1

2
3
4
1 A Review of SWAT Applications, Performance and Future Needs
5
6 2 for Simulation of Hydro-Climatic Extremes
7
8 3
9
10 4 Mou Leong Tan1*, Philip Gassman2, Xiaoying Yang3,4, James Haywood5
11
12 5
13
14 6 1Geography Section, School of Humanities, Universiti Sains Malaysia, 11800 Penang, Malaysia
15
16 7 2Center for Agricultural and Rural Development, Iowa State University, Ames, IA 50011-1054, USA
17
18 8 3Department of Environmental Science and Engineering, Fudan University, Shanghai 200433, China
19
20
9 4State Key Laboratory of Hydrology - Water Resources and Hydraulic Engineering, Nanjing Hydraulic
21
10 Research Institute, Nanjing 210029, China.
22
23
11 5College of Engineering, Mathematics and Physical Sciences, University of Exeter, United Kingdom
24
25
26
12 Abstract
27
28
13 Hydro-climatic extremes, such as droughts and floods, have most likely increased due to climatic
29
14 change and could lead to severe impacts on socio-economic, structural and environmental sectors. With
30
15 nearly 4,000 publications, the Soil and Water Assessment Tool (SWAT) is clearly one of the most
31 16 extensively used ecohydrological models worldwide. The model has been widely used for projecting
32 17 the impacts of future hydro-climatic changes, but application for extreme streamflow conditions is still
33 18 rarely reported. To date, SWAT application reviews have focused on compilations of SWAT studies
34 19 for specific or relatively new applications such as eco-hydrological modelling, ecosystem services, sub-
35 20 daily simulations, and pesticide fate and transport simulations. However, no existing SWAT review
36 21 studies have focused on simulation of hydro-climatic extremes. Therefore, this research aims to bridge
37 22 this gap by compiling and reviewing the findings of studies reporting SWAT hydro-climatic extremes
38 23 including highlighting the performance and future research needs. A total of 111 articles have been
39 24 identified since 1999; most of these studies were conducted in the United States and China. These
40 25 articles can be divided into extreme flow assessments, drought studies, flood studies, drought and flood
41 26 studies, SWAT coupling with other models, and SWAT improvements. Most of the extreme
42 27 performance assessment studies reported “satisfactory” performance, with a particular emphasis on
43 28 peak flow comparisons. Future research needs regarding this topic include: (1) a unified SWAT extreme
44 29 performance assessment framework; (2) SWAT improvements that result in improved replication of
45 30 peak and low flows; (3) reliability assessment of global and satellite products for SWAT extreme
46 31 simulations; (4) bias correction of CMIP6 and regional climate projections; (5) comparison of SWAT+
47 32 and SWAT for extreme flows simulations in different types of basins; (6) development of an extreme
48 33 flow module within an overall SWAT modelling system; and (7) integration of artificial intelligence
49 34 within SWAT modelling.
50 35
51
52 36 Keywords: SWAT; Flood; Drought; Extreme; Peak Flow; Climate Change
53
54 37
55
56 38
57
58 1
59
60
61
62
63
39 1.0 Introduction
64
65
40 Hydro-climatic extremes, such as droughts and floods, have most likely increased due to
66
41 climate change and could lead to severe impacts on socio-economic sectors (Giorgi et al., 2018, Raikes
67
68
42 et al., 2019). Concern about hydro-climatic extremes and their consequences has been rising in recent
69
70
43 years, as evidenced by the large effort of the Intergovernmental Panel on Climate Change (IPCC) in
71
72
44 summarising climate extreme impacts on droughts (sub-chapter 3.1) and floods (sub-chapter 3.2) in the
73
74 45 Special Report on Extremes (SREX) (IPCC, 2012). The report stated that some regions have
75
76 46 experienced increasing length and intensity of flood and/or drought events. By referring to the definition
77
78 47 of “climate extreme” in the SREX, the present study defines hydro-climatic extreme as “the occurrence
79
80 48 of a hydro-climatic variable’s value higher (or lower) than a defined extremely high (or low) threshold
81
82 49 value”.
83
84 50 Quantification of hydro-climatic extremes benefits local authorities and researchers in
85
86 51 understanding the current status and potential risks of water related disasters. It is advantageous to use
87
88 52 long-term observation data to measure the trends of hydro-climatic extremes for a specific region,
89
90 53 because it is the most reliable source for understanding the hydro-climatic system (Fu et al., 2010).
91
92 54 However, in many regions, the availability of reliable long-term observations remains a strong
93
94 55 limitation (Mishra and Singh, 2011, Richts and Vrba, 2016, Ummenhofer and Meehl, 2017). Moreover,
95
96 56 reliable long-term hydrological data are normally even less available as compared to climate data.
97
98 57 Therefore, provided the models are accurate, ecohydrological modelling could be used to simulate and
99
100 58 understand the hydro-climatic extremes and associated environmental impacts in regions without
101
102
59 reliable hydrological observations.
103
60 Ecohydrological modelling is an indispensable tool in understandig the interaction of
104
105
61 hydrological processes and environmental issues. Major advantages of ecohydrological modelling
106
107
62 include applicability for a wide range of basin scales and environment conditions and their capability
108
109 63 to perform cost effectiveness assessments and simulate “what-if” scenarios for planning purposes.
110
111 64 Simulated streamflow could be used to further calculate hydrological and pollutant extremes at
112
113 65 locations without a streamflow gauge. Moreover, ecohydrological models can be used to understand
114
115
116
117 2
118
119
120
121
66 the effectiveness of mitigation strategies in hydro-climatic events. An extensive array of hydrological
122
123
67 and ecohydrological models have been developed, many of which are described in previous review
124
125 68 studies (Shepherd et al., 1999, Singh and Woolhiser, 2002, Borah and Bera, 2004, Daniel, 2011, Fu et
126
127 69 al., 2019).
128
129 70 The Soil and Water Assessment Tool (SWAT) ecohydrological model (Arnold et al., 1998,
130
131 71 Arnold et al., 2012a, Williams et al., 2008, Bieger et al., 2017) is one of the most widely used models
132
133 72 worldwide and has been applied for an extensive suite of water resource issues across a broad range of
134
135 73 basin scales and environmental conditions, as evidenced by approximately 4,000 documented
136
137 74 publications (CARD, 2019). To date, several SWAT review studies have been reported in the literature
138
139 75 which can be divided into four main categories: (1) general review, (2) special issue review, (3) specific
140
141 76 region review, and (4) specific application review. General reviews include broad overviews of SWAT
142
143 77 applications, performance and future research needs (Gassman et al., 2007) as well as descriptions of
144
145 78 specific versions of SWAT (Bieger et al., 2017, Arnold et al., 1998). Special issue reviews provide a
146
147 79 summary of the topics and findings of multiple SWAT studies that were published in a special issue of
148
149 80 a specific journal; e.g. Krysanova and White (2015), Gassman et al. (2014), Douglas-Mankin et al.
150
151 81 (2010), Gassman and Wang (2015), Tuppad et al. (2011). As SWAT is increasingly applied across the
152
153 82 globe, compilations of SWAT study findings could be useful for SWAT developers or new users to
154
83 identify its major applications, capabilities, challenges and limitations in a specific region. To date,
155
156
84 regional-based SWAT reviews have been conducted for the Upper Nile River Basin region (van
157
158
85 Griensven et al., 2012), Brazil (Bressiani et al., 2015) and Southeast Asia (Tan et al., 2019a). Lastly, a
159
160
86 specific application review refers to a compilation of SWAT results and issues for studies focused on
161
162 87 specific or relatively new applications such as eco-hydrological modelling (Krysanova and Arnold,
163
164 88 2008), ecosystem services (Francesconi et al., 2016), pesticide fate and transport (Wang et al., 2019)
165
166 89 and sub-daily applications (Brighenti et al., 2019).
167
168 90 One of the earliest SWAT hydro-climatic extreme studies was conducted by Van Liew et al.
169
170 91 (2003) to understand how flood retention structures would affect streamflow characteristics during dry,
171
172 92 average and wet climate conditions. Narasimhan and Srinivasan (2005) and Hwang et al. (2006) used
173
174 93 SWAT-simulated soil moisture and evapotranspiration to measure drought indices for studying drought
175
176 3
177
178
179
180
94 patterns. Similar studies have been increasingly conducted in different regions of the world with more
181
182
95 specific focus on extreme hydrological conditions, including coupling models and developing extreme
183
184 96 indices as part of the analyses. However, a review of studies that report application of SWAT for the
185
186 97 assessment of hydro-climatic extremes has not been previously reported; therefore, this study was
187
188 98 conducted to fill this research gap. The overall aim of this study is to summarize the findings of existing
189
190 99 studies of SWAT for the assessment of hydro-climatic extremes for answering common questions that
191
192 100 might arise before, during and after the SWAT simulation as shown in Figure 1. The scope of this
193
194 101 review is limited to hydrologic assessment only, due to the need to provide an in-depth analysis of
195
196 102 SWAT’s ability to replicate extreme hydrological events. However, the insights gained from this
197
198 103 research are foundational for any SWAT application and have important implications regarding use of
199
200 104 the model for testing or scenarios that include simulation of pollutant transport.
201
202 105
203
204
205 106 2.0 SWAT Model Description, Selected Articles and Statistical Evaluation
206
207 107 2.1 SWAT
208
209 108 Current versions of SWAT feature a legacy of nearly continuous model development over
210
211 109 multiple decades as described in several previous studies (Arnold et al., 1998, Arnold et al., 2012a,
212
213 110 Gassman et al., 2007, Williams et al., 2008, Bieger et al., 2017). SWAT is comprised of several different
214
215 111 components including climatic inputs, crop growth and yield, hydrological cycling, representation of
216
217 112 management practices, erosion processes and resulting sediment transport, and pollutant (nutrient,
218
219 113 pesticide and pathogen) cycling and transport. The model is usually executed using a daily time step
220
221 114 although options are also provided to apply it using a sub-daily time step (Arnold et al., 2012b).
222
223 115 Partitioning of precipitation at the soil surface between runoff and infiltration, for daily time step
224
225 116 simulations, is performed using one of three variants of the U.S. Department of Agriculture (USDA)
226
227 117 Natural Resources Conservation Service (NRCS) runoff curve number (RCN) method (Arnold et al.,
228
229 118 2012b, USDA-NRCS, 2004a, USDA-NRCS, 2004b, Williams et al., 2012). The Green and Ampt Mein
230
119 Larson (GAML) excess rainfall method (Mein and Larson, 1973), an adaption of the original Green and
231
232
120 Ampt infiltration method (Green and Ampt, 1911), is used to partition precipitation between surface
233
234
235 4
236
237
238
239
121 runoff and infiltration for sub-daily time step applications (Jeong et al., 2010, Arnold et al., 2012b,
240
241
122 Brighenti et al., 2019).
242
243 123 Basins simulated in SWAT are first delineated into sub-basins, which are then sub-divided into
244
245 124 hydrologic response units (HRUs) which represent landscapes consisting of homogeneous soil,
246
247 125 topographic, land use and management characteristics. These HRUs represent percentage land areas
248
249 126 within a sub-basin and are not spatially recognized by SWAT version 2012 (Gassman and Wang, 2015).
250
251 127 Enhanced spatial representation of cropland landscapes and other features are possible in newer
252
253 128 SWAT+ codes, although HRUs are primarily configured within larger Landscape Units as described by
254
255 129 (Bieger et al., 2017), which represent the latest version of SWAT that was released in 2019 (SWAT,
256
257 130 2019c). Landscape-level hydrology flows and pollutant losses are estimated at the HRU level, which
258
259 131 are then summed to the corresponding sub-basins outlet and routed via the stream network to the overall
260
261 132 basin outlet. Further details regarding the required inputs for SWAT and the range of available outputs
262
263 133 are provided in Arnold et al. (2012b).
264
265 134
266
267 135 2.2 Article selection process
268 136 The SWAT Literature Database (CARD, 2019) and Web of Science (WoS) database were used
269
270 137 to identify SWAT hydro-climatic extremes related research articles. Four main criteria were used for
271
272 138 the selection of articles. First, the article title should contain the keywords “extreme”, “flood”,
273
274 139 “drought”, “peak flow”, “high flow” and/or “low flow.” This criterion was adopted to ensure that the
275
276 140 selected studies specifically focused on the topic. Second, only articles that were published from 1998
277
278 141 on were considered, because that was the year that the first major description of SWAT was published
279
280 142 (Arnold et al., 1998) and also marked the beginning of when the extensive SWAT literature began to
281
282 143 expand (CARD, 2019). Third, peer-reviewed articles that were included in WoS Core Collection (CC)-
283
284
144 indexed journals (Analytics, 2019) were the primary source of information considered in this study.
285
145 Grey literature such as government reports, technical reports, conference papers and non-English
286
287
146 articles were thus excluded from this review. Finally, the selection of articles was also based on the
288
289
147 authors’ judgement who have considerable knowledge of the existing SWAT-related literature. For
290
291 148 example, some articles were included even though their keywords did not match the searching criteria.
292
293
294 5
295
296
297
298
149 Searching priority was given to the 3,821 articles contained in the SWAT literature database
299
300
150 (as of July 15, 2019) because the articles have been filtered and grouped according to specific categories
301
302 151 (CARD, 2019). Then, the article selection was based on the four criteria that were mentioned earlier
303
304 152 including the WoS CC, which contains additional up-to-date articles that have not been entered in the
305
306 153 SWAT Literature Database yet. A total of 147 articles were initially identified in the WoS CC (again
307
308 154 on July 15, 2019), based on one or more of the previous keywords within the article title in combination
309
310 155 with the additional term “SWAT” as a topic in the WoS. Unrelated studies were then excluded from the
311
312 156 review; e.g. flood modelling for the “SWAT” River, Pakistan. In total, 111 articles related to SWAT
313
314 157 hydro-climatic extreme simulations were ultimately identified (Table 1) based on the two searches.
315
316 158 The selected articles were further analysed in terms of region, publication year and basin size
317
318 159 (Figure 2). Most of the SWAT hydro-climatic studies were conducted in the United States (34.51%)
319
320 160 and China (13.27%), as shown in Figure 2c. The reason for the high rate of application in the United
321
322 161 States is that the SWAT model was originally developed using United States environmental, land and
323
324 162 soil conditions (Arnold et al., 1998). Moreover, a series of SWAT workshops and hands-on training
325
326 163 organized by Texas A&M AgriLife has been a further catalyst for increased studies in the United States.
327
328 164 Similar studies have also been conducted in transboundary basins (10.62%) such as the Brahmaputra
329
330 165 River Basin (Mohammed et al., 2017b) and Mekong River Basin (Arias et al., 2014), showing that
331
166 understanding of extreme changes is very important for transnational basin management.
332
333
167 Publication of SWAT studies has increased dramatically since 2009 (Tan et al., 2019a, CARD,
334
335
168 2019), but SWAT assessment of hydro-climatic extremes has only started to increase rapidly from 2017.
336
337
169 Figure 2a indicates that the number of articles published in 2017 was three times greater than those
338
339 170 published in 2016. A possible explanation for this might be the increase of flood and drought disasters
340
341 171 around the world in the past few years (Chou et al., 2013, Naveendrakumar et al., 2019). Increasing
342
343 172 awareness of climate change and its impact on regional hydrological processes has resulted in the urgent
344
345 173 formulation of hydro-climatic related policies based on the possible effects of future extreme events.
346
347 174 SWAT has been utilized for future assessment of hydro-climatic extremes in dozens of studies (CARD,
348
349 175 2019) because it can simulate “what-if” scenarios easily. Another reason for the recent increase in
350
351 176 SWAT applications for extremes applications could be the availability of hydro-climatic related
352
353 6
354
355
356
357
177 international grants. For example, an international research grant on “Understanding the Impacts of
358
359
178 Hydro-meteorological Hazards in Southeast Asia” was funded by the Newton fund, the UK’s Natural
360
361 179 Environment Research Council (NERC), the Economic and Social Research Council (ESRC) and local
362
363 180 governments. Such funding could result in increased numbers of SWAT hydro-climatic studies in
364
365 181 Southeast Asia (Netwon-NERC, 2019).
366
367 182 Figure 2b indicates the studies that report application of SWAT for the assessment of hydro-
368
369 183 climatic extremes were mainly conducted in basins of 10,001-100,000 km2 (34.19%) in size, followed
370
371 184 by >100,000 km2 (17.95%), 101-1,000 km2 (16.24%) and 1,001-5,000 km2 (15.38%). By contrast, only
372
373 185 5.13% of the studies were conducted in basins that were smaller than 100 km2 (Figure 2b). These trends
374
375 186 underscore that SWAT hydro-climatic extreme studies were mainly applied in basins of medium to
376
377 187 large scales. The biggest study area was the Zambezi River Basin with a drainage area of 1,400,000
378
379 188 km2, which drains parts of Angola, Botswana, Malawi, Mozambique, Namibia, Tanzania, Zambia and
380
381 189 Zimbabwe. The result shows that more studies should be conducted in smaller scale basins in the future.
382
383 190
384
385 191 2.3 Evaluation of the accuracy of SWAT output
386 192 Explicit standards have not been established for evaluating the accuracy of hydrological and
387
388 193 eco-hydrological models, although various protocols have been proposed for judging simulated
389
390 194 hydrological and/or pollutant outputs (Engel et al., 2007, Arnold et al., 2012a, Nair, 2011, Harmel et
391
392 195 al., 2018). Typically, both graphical and statistical measures are used to judge model accuracy (Arnold
393
394 196 et al., 2012a). Coffey et al. (2004) describe over a dozen different statistical measures that can be used
395
396 197 to evaluate hydrological model output including the root mean square error (RMSE), coefficient of
397
398 198 determination (R2) and Nash-Sutcliffe Efficiency (NSE). Krause et al. (2005) provide further insights
399
400 199 regarding the strengths and weaknesses of the R2 and NSE statistics. Other statistical measures have
401
402
200 been introduced more recently such as the Kling-Gupta Efficiency (KGE), which is described by Guse
403
201 et al. (2017). To date, the R2 and NSE statistics remain the dominant measures that have been used to
404
405
202 assess the accuracy of SWAT model output (Gassman et al., 2007, Gassman et al., 2014, Tuppad et al.,
406
407
203 2011, Bressiani et al., 2015, Tan et al., 2019a). Criteria suggested by Moriasi et al. (2007), Moriasi et
408
409
410
411
412 7
413
414
415
416
204 al. (2015) are frequently used to interpret whether calculated NSE and R2 values reflect satisfactory
417
418
205 model results, and are used in this study to discuss the accuracy of reported model outcomes.
419
420 206 The distribution of R2 and NSE statistics that were reported among the 111 articles reviewed
421
422 207 for this study is shown in Figure 4. These statistics are based on comparisons of SWAT simulated
423
424 208 streamflow values versus corresponding measured values for monthly (aggregated over daily values),
425
426 209 daily and sub-daily time periods. Nearly 90% of the NSE statistics and over 90% of the R2 statistics
427
428 210 exceed 0.5 and 0.6, respectively, which are the thresholds suggested by Moriasi et al. (2007) and/or
429
430 211 Moriasi et al. (2015) for satisfactory simulation results. Many of the computed statistics would further
431
432 212 satisfy criteria of “good” and “very good” streamflow simulation results as proposed by Moriasi et al.
433
434 213 (2007). The relative distribution of the monthly and daily statistics shown in Figure 4 is consistent with
435
436 214 previous similar distributions reported by Tuppad et al. (2011), Gassman et al. (2014) and Tan et al.
437
438 215 (2019a). The distribution of sub-daily statistics (Figure 4) is similar to the graphical summaries of NSE
439
440 216 statistics reported by Brighenti et al. (2019) as part of their review of studies that focused on SWAT
441
442 217 sub-daily time step applications.
443
444 218
445
446
447 219 3.0 Application
448
449 220 The 111 articles were further divided into six major SWAT categories based on their primary
450
451 221 focus: (1) SWAT performance regarding replication of extreme flows, (2) SWAT drought-related
452
453
222 studies, (3) SWAT flood-related studies, (4) SWAT studies that incorporate both drought and flood
454
223 analyses, (5) SWAT coupling with other models, and (6) SWAT model applications featuring
455
456
224 modifications, enhanced pre- or post-processing capabilities and/or some other improvement (Table 1).
457
458
225 The rationale of these categories is to better describe the major findings that are reported in each article.
459
460
226 Some articles might fall under two or three categories, but the first priority was given to the “SWAT
461
462 227 performance regarding replication of extreme flows” category. This is because the reliability of SWAT
463
464 228 to replicate extreme flows is among the major concerns of SWAT users before applying the model for
465
466 229 hydro-climatic extremes studies. Also, the majority of existing studies did not evaluate the ability of
467
468 230 SWAT to replicate extreme flows.
469
470
471 8
472
473
474
475
476
231 3.1 SWAT performance regarding simulation of extreme flows
477
478
232 This section addresses questions related to SWAT’s reliability regarding replication of extreme
479
233 events such as: (1) “How to calibrate and validate SWAT for simulating extreme events?”, and (2)
480
481
234 “How accurately does SWAT replicate extreme events?” (Figure 1). SWAT is a continuous-time model,
482
483
235 and thus hydrologic calibration and validation are mainly based on continuous observed streamflow
484
485
236 data. Performance of SWAT for continuous simulations has been well-proven in many regions and
486
487 237 documented in Figure 4 and in previous SWAT reviews (Tan et al., 2019a, Bressiani et al., 2015,
488
489 238 Gassman et al., 2014, Tuppad et al., 2011). Besides that, a comprehensive review covering SWAT sub-
490
491 239 daily scale simulations has recently been conducted by Brighenti et al. (2019). Therefore, this review
492
493 240 only focuses on SWAT performance assessments that report further validation in terms of specific flood
494
495 241 events (event-based), peak flow, low flow, different precipitation intensities and/or flow duration curve
496
497 242 simulations, which were analysed beyond typical continuous-time assessment.
498
499 243 Most of the studies only considered one of the criteria (specific flood event, peak flow, low
500
501 244 flow, different precipitation intensities or flow duration curve) for their respective SWAT extreme
502
503 245 assessment. For example, Bacopoulos et al. (2017) calibrated and validated SWAT using an hourly time
504
505 246 step for specific flood events for the St. Johns River in the United States. They concluded that SWAT
506
507 247 could reasonably capture flood events, based on NSE values of 0.85 and 0.45 computed for the
508
509 248 simulated calibration and validation periods, respectively. Campbell et al. (2018) calibrated and
510
511 249 validated SWAT for Pawtuxet River Basin in the United States, using a daily time step for respective
512
513 250 one-year periods in 2010 and 2013 that resulted in NSE values ranging from 0.62 to 0.69. They then
514
515
251 assessed the effects of urbanization on flood events with SWAT using a sub-daily simulation time step
516
252 of 5 minutes. Pfannerstill et al. (2014) is among the limited studies that evaluated the ability of SWAT
517
518
253 to reproduce both extreme high and low flows. They developed a multi-metric performance framework
519
520
254 by dividing a flow duration curve into five different segments: very low, low, medium, high and very
521
522 255 high flows. They concluded that the use of this performance framework could result in improved
523
524 256 calibration of SWAT, which in turn could lead to better prediction of extreme flows.
525
526
527
528
529
530 9
531
532
533
534
257 Boithias et al. (2017) and Yu et al. (2018a) merged both continuous-time and event-based
535
536
258 assessments in their respective studies. The latter authors developed a SWAT-Event model that refined
537
538 259 the lumped unit hydrograph from the original SWAT into a set of distributed units to represent spatial
539
540 260 variability of a basin. The model was able to accurately reproduce 24 flood events that occurred in the
541
542 261 Wangjiaba River, China, as evidenced by NSE values as high as 0.95. Li et al. (2018a) incorporated a
543
544 262 flow duration curve assessment as part of their overall evaluation approach that included the continuous-
545
546 263 time and event-based approaches mentioned earlier. Meanwhile, Yaduvanshi et al. (2018) introduced
547
548 264 precipitation intensity assessment to better understand the performance of SWAT under low, moderate
549
550 265 or extreme precipitation events. Both studies reported that SWAT performed well in the context of flood
551
552 266 simulations. A combination of continuous-time and peak flows comparison assessment is among the
553
554 267 typical approaches used to evaluate peak flows (Dakhlalla and Parajuli, 2016, Javaheri and Babbar-
555
556 268 Sebens, 2014, Spellman et al., 2018, Zhang et al., 2015). Javaheri and Babbar-Sebens (2014) estimated
557
558 269 simulated sub-daily peak flows for the Eagle Creek basin located in central Indiana in the United States,
559
560 270 based on an relationship between the continuous-time daily flows and the sub-daily flows. They found
561
562 271 the errors of the predicted SWAT-based peak flows varied only from 1 to 18% relative to the
563
564 272 corresponding observed peak flows.
565
566 273 Due to the development of satellite technologies, Kumar and Lakshmi (2018) and Tan et al.
567
274 (2018) studied the performance of satellite precipitation products in relation to SWAT estimates of
568
569
275 extreme events for the Gandak River Basin (India, China and Nepal) and the Kelantan River Basin
570
571
276 (Malaysia), respectively. Kumar and Lakshmi (2018) evaluated the performance of the Tropical
572
573
277 Rainfall Measuring Mission (TRMM) (Huffman et al., 2007) product using a continuous-time
574
575 278 simulation coupled with varying precipitation intensities (light, moderate, heavy and extremely heavy
576
577 279 precipitation). They found that the TRMM-based SWAT model performed better for the moderate and
578
579 280 heavy precipitation periods versus the light and extremely heavy precipitation periods. In contrast, Tan
580
581 281 et al. (2018) used a combined continuous-time and event-based assessment to test the reliability of
582
583 282 driving SWAT with three Global Precipitation Mission (GPM) products (Hou et al., 2014). A
584
585 283 “satisfactory” performance was reported for SWAT in replicating the “Big Yellow Flood” event using
586
587 284 the GPM products.
588
589 10
590
591
592
593
594
285 3.2 SWAT studies on drought
595
596
286 Understanding of historical and future droughts in terms of intensity, duration, severity and
597
287 spatial extent is very important for freshwater management and planning (Mishra and Singh, 2011). For
598
599
288 example, priority irrigation systems could be built for sub-basins with high drought risk to reduce
600
601
289 agricultural losses. Drought can be mainly divided into meteorological, hydrological, agricultural and
602
603
290 socio-economic droughts (Wilhite and Glantz, 1985, Mishra and Singh, 2011). SWAT is commonly
604
605 291 used to evaluate the impacts of meteorological and hydrological droughts. In general, SWAT-based
606
607 292 drought studies can be divided into two main groups: (1) index-based, which are based on calculation
608
609 293 and analysis of drought indices from SWAT outputs, and (2) non index-based, which involve analysis
610
611 294 of SWAT-simulated streamflow in specific low flow periods. In general, future climate projections are
612
613 295 incorporated into a calibrated SWAT model to assess the effects of predicted future drought processes.
614
615 296 The projected future SWAT outputs are then compared with historical values using various drought
616
617 297 indices. The comparison could be done by considering the potential drought risks such as intensity,
618
619 298 occurrences, severity and duration in the future.
620
621 299 A drought index is a simplified variable representation of drought severity that uses hydro-
622
623 300 climatic inputs such as precipitation, temperature and streamflow. Information about commonly used
624
625 301 drought indices, including their strengths and weaknesses is available in a drought indices handbook
626
627 302 published by the World Meteorological Organization (WMO) and Global Water Partnership (GWP)
628
629 303 (2016). Standardized Precipitation Index (SPI), Standardized Streamflow Index (SSI), Standardized
630
631 304 Soil Moisture Index (SSMI) and Palmer Drought Severity Index (PDSI) are popular indices that are
632
633
305 frequently used in SWAT drought studies. Tan et al. (2019b), Vu et al. (2015) and Zhao et al. (2019)
634
306 incorporated climate projections into the SWAT model to calculate future SPI and SSI for river basins
635
636
307 in Malaysia, Vietnam and China, respectively. Meanwhile, the SSMI was considered and measured by
637
638
308 Narasimhan and Srinivasan (2005), Kamali et al. (2017) and Li et al. (2017) using SWAT-simulated
639
640 309 soil moisture data. Kang and Sridhar (2017b) and Zou et al. (2017) modified the PDSI to study drought
641
642 310 impact in the United States and China, respectively, using streamflow, soil moisture and
643
644
645
646
647
648 11
649
650
651
652
311 evapotranspiration data within SWAT. Most of the studies reported that more severe drought conditions
653
654
312 might occur in the future.
655
656 313 Some SWAT drought studies concentrated on analysing potential future low flows. Rahman
657
658 314 et al. (2010) evaluated the climate change impact on low flows of the Ruscom River Basin, located in
659
660 315 southern Ontario in the eastern part of Canada. They found a potential reduction of up to 50% in the
661
662 316 annual minimum monthly flow of a five-year return period. They also reported that the low flows might
663
664 317 decrease in the summer and fall, but increase in the spring. The lowest 7-day low flow of each year with
665
666 318 a 10-year return period (7Q10) flow was also considered in SWAT analyses conducted in South Korea
667
668 319 (Ryu et al., 2011) and the United States (Shrestha et al., 2017, Shrestha et al., 2019). Ryu et al. (2011)
669
670 320 found a significant reduction of the 7Q10 flow from 1.54 m3s-1 to 0.03 m3s-1 under the most severe
671
672 321 scenario on the Geum River Basin in South Korea. On the other hand, a flow duration curve (FDC)
673
674 322 calculated from SWAT outputs has also been applied in drought analysis (Brown et al., 2015, Hoyos et
675
676 323 al., 2019, Rahman et al., 2010). The 75 (Q75) and 95 (Q95) percentiles are normally considered as the
677
678 324 threshold for extreme low flow and low flow, respectively (Shrestha et al., 2017). Li et al. (2018b)
679
680 325 applied the 80 (Q80) percentile to precipitation (station), streamflow (SWAT) and soil moisture (SWAT)
681
682 326 for comparison of meteorological, hydrological and agricultural droughts over the Luanhe River Basin,
683
684 327 China. The modelling results showed that the number of hydrological and agricultural droughts were
685
328 less than meteorological droughts, but the durations was longer.
686
687
329
688
689
690 330 3.3 SWAT flood studies
691
692 331 In recent decades, floods have become more frequent and intense around the world.
693
694 332 Quantification of future flood risk is vital to reduce damage to infrastructure and human lives (Cloke
695
696 333 and Pappenberger, 2009) Although SWAT is not designed for flood modelling, there still have been
697
334 reported SWAT flood studies. Overall, SWAT was used to understand the impact of land use and/or
698
699
335 climate changes on floods at the basin scale. Besides that, the effectiveness of flood mitigation or
700
701
336 retention strategies can also be simulated easily with SWAT. To date, it has been calibrated and
702
703
337 validated at the daily scale in most of the SWAT flood-related studies (Maghsood et al., 2019, Cheng
704
705
706
707 12
708
709
710
711
338 et al., 2017, Mohammed et al., 2017a) versus monthly-scale testing that is normally reported in SWAT
712
713
339 drought studies (Tan et al., 2019b, Bayissa et al., 2018). Further subdivision of the SWAT-based flood
714
715 340 studies (Table 1) results in three main categories, based on the type of analyses: land use change, climate
716
717 341 change, and land use and climate change.
718
719 342 Assessment of the impact of land use management on flood risk using the SWAT model has
720
721 343 been conducted in China (Zhang et al., 2016), Spain (Jodar-Abellan et al., 2019), South Korea (Lee et
722
723 344 al., 2017), and the United States (Mitchell et al., 2018, Schilling et al., 2014, Van Liew et al., 2003).
724
725 345 Besides that, Angelidis et al. (2010) compared the flood hydrographs that were simulated by the
726
727 346 EvroFloods model they developed versus SWAT for the Evros/Maritza River Basin that drains portions
728
729 347 of Bulgaria, Greece and Turkey. They found an approximately 5.8% difference between the historical
730
731 348 and future peak flows. Mitchell et al. (2018) proved that water retention sites (WRS) could effectively
732
733 349 reduce the high flows of a Minnesota River sub-basin using the SWAT model. Schilling et al. (2014)
734
735 350 found a conversion of 50% to 100% of the cropland to perennial vegetation could dramatically reduce
736
737 351 the flood risk in the Raccoon River basin in west central Iowa.
738
739 352 SWAT-based climate change impact assessments are conducted by comparing the impacts of
740
741 353 historical climate patterns with projected future climate inputs within a calibrated SWAT model. The
742
743 354 future “climate input” is usually extracted from the Coupled Model Intercomparison Project (CMIP) 3
744
355 (Meehl et al., 2007) and CMIP5 (Taylor et al., 2012) General Circulation Models (GCMs) (Maghsood
745
746
356 et al., 2019, Iqbal et al., 2018, Kharel and Kirilenko, 2018) or based on potential “synthetic”
747
748
357 precipitation and temperature changes (Gao et al., 2018, Kehew et al., 2010). Kharel and Kirilenko
749
750
358 (2018) found the overspill risk ranged from 7.3% to 47.1% for Devils Lake in North Dakota, U.S. in
751
752 359 response to CMIP3 and CMIP5 climate projections. Maghsood et al. (2019) evaluated the flood impact
753
754 360 using the Flood Frequency Index (FFI) and Sub-basin Flood Source Area Index (SFSAI) as measured
755
756 361 from SWAT outputs in the Talar River Basin in northern Iran. Their findings showed that high potential
757
758 362 flood risk is found in sub-basins mainly located in the eastern part of the basin. Iqbal et al. (2018) used
759
760 363 HEC-SSP, a statistical software packages developed by US Army Corps of Engineers to conduct flood
761
762 364 frequency analysis as a function of SWAT-simulated streamflow for the Kabul River Basin that drains
763
764
765
766 13
767
768
769
770
365 parts of Afghanistan and Pakistan,. They found that a 50-year return period flood is likely to occur more
771
772
366 frequently in the future.
773
774 367 Assessing combined land use change and climate change impacts could provide a more robust
775
776 368 flood risk assessment (Huang et al., 2018). Igarashi et al. (2019) incorporated synthetic climate
777
778 369 projections and future projected land use changes from the Conversion of Land Use and its Effects at
779
780 370 Small regional extent (CLUE-S) model (Verburg et al., 2002) into SWAT to study high flows in the
781
782 371 Song Khwae district, Thailand. The modelling showed that the high flows are projected to decline in
783
784 372 the future even with an increase of cropland up to 50% within the basin. Cheng et al. (2017) proposed
785
786 373 the need for flood retention sites to reduce potential future flooding based on SWAT simulations driven
787
788 374 by low and moderate emission scenarios climate projections in Charles River Basin, which drains
789
790 375 portions of the Boston metropolitan area in eastern Massachusetts, U.S.. Similarly, Walters and Babbar-
791
792 376 Sebens (2016) applied climate projections from the North American Regional Climate Change
793
794 377 Assessment Program (NARCCAP) (Mearns et al., 2009) in SWAT to study the effectiveness of
795
796 378 wetlands in mitigating high flows in the Eagle Creek Basin, U.S.. They concluded that the wetlands
797
798 379 could reduce the intensity of peak flows by 15% to 20%.
799
800 380
801
802 381 3.4 SWAT studies on drought and flood
803
804 382 Consideration of both low and peak flows could provide an overview of hydro-climatic extreme
805
806 383 changes and help to identify which natural hazard will have more impact in the future. Local authorities
807
808 384 can thereafter allocate more funding and expertise to mitigate the specific hazard. Interestingly, the
809
810 385 Coordinated Regional Climate Downscaling Experiment (CORDEX) (Giorgi et al., 2009) has proven
811
812 386 to be one of the most popular climate projections applied for this topic. The CORDEX – South Asia
813
814 387 domain database with a spatial resolution of 0.5o (~50 km) was incorporated into SWAT to study low
815
388 and peak flows for the Brahmaputra River system in south Asia (Mohammed et al., 2017b, Mohammed
816
817
389 et al., 2017a). The modelling results showed that floods are projected to become more frequent and
818
819
390 intense compared to droughts in the future. Tirupathi et al. (2018) also incorporated the CORDEX South
820
821
391 Asia domains within SWAT to investigate the impacts of future climate projections on monsoonal
822
823
824
825 14
826
827
828
829
392 rainfall patterns and excessive precipitation events. Similarly, the CORDEX Africa domain was used
830
831
393 by Näschen et al. (2018) to assess the effects of future climate on wetland resources for the Kilombero
832
833 394 River Basin in Tanzania.
834
835 395 Several innovative computer tools have been developed to transfer SWAT and other hydrologic
836
837 396 output into useful hydrological extremes information. The Water Engineering Time Series PROcessing
838
839 397 tool (WETSPRO) was used by Leta and Bauwens (2018) and Leta et al. (2018) to extract daily extreme
840
841 398 low and peak flows from the SWAT-simulated daily streamflow in Belgium and Hawaii, respectively.
842
843 399 The tool is capable of selecting low and peak flows using the independent peak over threshold (POT)
844
845 400 method. By contrast, Chen et al. (2019a) measured hydrological extremes using the Indicators of
846
847 401 Hydrologic Alteration (IHA) tool (Richter et al., 1996). The IHA provides 67 statistical parameters
848
849 402 related to river ecosystems, including several parameters related to hydrological extremes (i.e. annual
850
851 403 1, 3, 7, 30, 90-day max or min flows, frequency and duration flood or extreme low flows) that can be
852
853 404 selected by users. These parameters could be measured using SWAT-simulated historical and future
854
855 405 streamflow projections, followed by a comparison between the two periods with statistical analysis.
856
857 406 Stewart et al. (2015) compared historical monthly low and peak flows from 1961 to 1990 with
858
859 407 a future period (2071 - 2099) using SWAT to understand their impact on the Sierra Nevada and Upper
860
861 408 Colorado River Basin (UCRB) eco-hydrological system in the U.S.. Their findings showed a significant
862
409 increase in the number of high flows in the winter and spring seasons, particularly in the UCRB. By
863
864
410 contrast, the extreme low flows of both basins are projected to increase in spring and summer. This
865
866
411 information is important to ecological scientists studying fish habitat and potential fish movement in
867
868
412 the future.
869
870 413
871
872
414 3.5 SWAT model coupling with other models
873
874
415 In some cases, a modeller might want to know the impact of hydro-climatic extremes on SWAT
875
876
416 outputs that are not based on flow characteristics, e.g. flood extent, aquatic life, agricultural yield,
877
878
417 economic losses and hydro-electricity capacity. Coupling of SWAT with other models helps to expand
879
880
418 the assessment to other parameters. In general, Hydrologic Engineering Center's River Analysis System
881
882
883
884 15
885
886
887
888
419 (HEC-RAS) (USACE, 2016), MIKE FLOOD (DHI, 2017) and SOBEK (Deltares, 2019) are some of
889
890
420 the most widely used flood models to study potential future flood patterns via couplings with SWAT
891
892 421 (Arunyanart et al., 2017, Song et al., 2014, Kuntiyawichai et al., 2011). For example, Robi et al. (2019)
893
894 422 coupled SWAT with MIKE FLOOD to identify the flood inundation extent and flooding depth under
895
896 423 CMIP5 climate projections for the Ribb River Basin which is located in northwest Ethiopia. They found
897
898 424 that the future flood extent area could cover up to 61.01 km2. Similarly, Chinnasamy et al. (2018)
899
900 425 coupled SWAT, MOD-FLOW (Waterloo, 2011) and HEC-RAS to test the potential of the Underground
901
902 426 of Taming of Floods for Irrigation (UTFI) system in reducing groundwater depletion and flood extent
903
904 427 in the Ramganga River Basin in India. The simulation results showed that the implementation of UTFI
905
906 428 can reduce the flood inundation areas by 10% and increase the groundwater level by 7 m.
907
908 429 An insurance fund simulator, called Hydrologic Risk Transfer Model (MTRH-SHS) (Righetto
909
910 430 and Mendiondo, 2007) has been developed by coupling SWAT with various vulnerability and financial
911
912 431 modules to explore drought and flood risks from an economic perspective (Mohor and Mendiondo,
913
914 432 2017). A case study of the Piracicaba River Basin in southeast Brazil showed that premiums up to 1%
915
916 433 of the local Gross Domestic Product (GDP) are required when water demand increases to 20%. Gies et
917
918 434 al. (2014) evaluated the impacts of three different drought adaptation policies in for Juba River in East
919
920 435 Africa using SWAT and a system dynamics model. They concluded a combination of hydraulic
921
436 infrastructure and new agricultural practices are the most effective policy to reduce drought impact,
922
923
437 based on guidance provided by the coupled modelling system. In Southeast Asia, Arias et al. (2014)
924
925
438 coupled SWAT, the Integrated Quantity and Quality Model (IQQM) (Simons et al., 1996), the
926
927
439 Hydrologic Engineering Center’s Reservoir System Simulation (HEC-ResSIM) (USACE, 2013) and
928
929 440 the two-dimensional Environmental Impact Assessment Model (2D-EIA) to quantify how hydropower
930
931 441 development could change the hydrological regime in the Sesan, Srepok, and Sekong (3S) river system.
932
933 442 They found that full damming will cause significant hydrological alternations, and that the local
934
935 443 authorities should practice sustainable hydropower practices in the future.
936
937 444
938
939
940
941
942
943 16
944
945
946
947
948
445 3.6 SWAT improvement
949
950 446 Efforts to improve SWAT to support more accurate hydro-climatic extreme simulations have
951
952 447 focused on development of a new drought or flood index, calibration method improvement, uncertainty
953
954 448 analysis and/or SWAT modification. A specific crop-based drought index calculated from precipitation,
955
449 temperature, soil moisture depletion, transpiration and biomass production, with the latter three
956
957
450 variables simulated by SWAT, was developed to quantify agricultural drought in the United States
958
959
451 (McDaniel et al., 2017a, McDaniel et al., 2017b, McDaniel et al., 2017c). Esfahanian et al. (2017)
960
961
452 developed a new drought index, called Meteorological, Agricultural, Stream Health and Hydrological
962
963 453 (MASH) that merges 13 drought indices. These newly developed indices could effectively identify the
964
965 454 risk of flood or drought.
966
967 455 Reliable model calibration is essential for providing a strong foundation to support scientific
968
969 456 analysis and decision making. Chilkoti et al. (2018) proposed a multi-objective auto-calibration method
970
971 457 for improving SWAT low flow simulations. The results showed that the volume efficiency and time
972
973 458 series of low flow simulations in the Saugeen River, Canada, were improved by 135% and 65%,
974
975 459 respectively. Trudel et al. (2017) evaluated low flow uncertainties based on the SWAT structure and
976
977 460 approach that was simulated for the Yamaska River Basin in Canada. They found that the calibration
978
979 461 objective function is one of the main uncertainty sources for SWAT low flow simulations. Another
980
981 462 SWAT extreme flow uncertainty study was conducted by Zhang et al. (2014) in the Qiantang River
982
983 463 Basin, China. The authors considered three major uncertainties that normally arise in SWAT for the
984
985 464 assessment of hydro-climatic extremes, namely climate emission scenarios (A1B, A2 and B2), extreme
986
987 465 values model (Generalized Pareto distribution, Generalized Extreme Value distribution and Pearson
988
989 466 Type 3 distribution) and SWAT parameters (sequential uncertainty method), in their extreme
990
991
467 uncertainty analysis. Interestingly, the SWAT model parameters showed a larger uncertainty than the
992
468 emission scenario and extreme value models for small return periods simulations. Meanwhile, the
993
994
469 uncertainties of the three sources for large return period simulations are likely to be similar.
995
996
470 Duan et al. (2018) modified the snow-melting and flood process modules to improve peak flow
997
998
471 simulation in SWAT, by incorporating the accumulated temperature and maximum temperature
999
1000
1001
1002 17
1003
1004
1005
1006
472 elements in a snow melt processes module within SWAT. They tested the modified SWAT for the
1007
1008
473 Tizinafu River basin in China, and found a 43% improvement in the representation of the observed peak
1009
1010 474 flows. Cohen Liechti et al. (2014) developed a SWAT reservoir model to better model flood plain
1011
1012 475 behaviour in the Zambezi Basin in the southern African continent. For instance, they modified the
1013
1014 476 outflow of the reservoir sub-model of the SWAT2009 model by separating overflow from base flow.
1015
1016 477 The findings indicated that the modified SWAT model resulted in superior replication of low and high
1017
1018 478 flows versus the original SWAT model. Other efforts have focused on modifying the RCN method or
1019
1020 479 other hydrological functions in SWAT, resulting in improved representation of peak flows and overall
1021
1022 480 replication of measured streamflows and other hydrological indicators (Xie et al., 2020, D. White et al.,
1023
1024 481 2009, Kim and Lee, 2008).
1025
1026 482
1027
1028
1029 483 4.0 Future Research Directions
1030
1031 484 4.1 SWAT low and high flows assessment framework
1032
1033 485 General procedures for building a SWAT model include: (1) creating the model with a GIS
1034
1035 486 interface; e.g. ArcGIS SWAT (ArcSWAT) (Olivera et al., 2006, SWAT, 2019a) or the QSWAT
1036
1037 487 interface (Dile et al., 2016, SWAT, 2019b), (2) parameter sensitivity analysis, (3) calibration and
1038
1039 488 uncertainty analyses, (4) validation of the calibrated model, and (5) performing an impact assessment.
1040
1041 489 Typically, observed streamflow data is divided into two time periods, with one-time period used for
1042
1043 490 sensitivity analysis and calibration while the other one is used for validation. This split-time calibration
1044
1045 491 and validation framework has been applied in the most of the SWAT hydro-climatic extreme studies
1046
1047 492 reviewed here. General guidelines, main issues and solutions for SWAT sensitivity analysis, calibration
1048
1049 493 and validation are available in many previous studies (Arnold et al., 2012a, Abbaspour et al., 2018,
1050
1051 494 Moriasi et al., 2007, Moriasi et al., 2015). Here, we only discuss the main issues and possible solutions
1052
1053 495 for simulating extreme events.
1054
1055 496 Figure 3 shows the percentage of the finest time-scale streamflow observations that were used
1056
497 in the selected publications. Daily streamflow was used for calibration & validation in nearly half of
1057
1058
498 the studies versus monthly streamflow in the calibration process for about 25% of the studies. Only
1059
1060
1061 18
1062
1063
1064
1065
499 about 5% of the studies calibrated their model with streamflow data at a sub-daily scale. In general,
1066
1067
500 monthly scale calibration and validation was used to support drought studies. This is because most of
1068
1069 501 the drought index calculations only need monthly data; e.g., Standardized Streamflow Index (SSI).
1070
1071 502 Meanwhile, SWAT-based flood studies were mainly calibrated using daily data (Maghsood et al., 2019,
1072
1073 503 Xu et al., 2017) although some required hourly-scale time steps to capture specific flood events.
1074
1075 504 Widespread availability of hourly streamflow observations and hourly climate data to drive the model
1076
1077 505 are still lacking, particularly in developing and less developed countries. However, daily streamflow
1078
1079 506 data are sufficient for most long-term flood and drought analyses, but a more comprehensive calibration
1080
1081 507 and validation framework is needed for these type of applications.
1082
1083 508 Generally, studies that relied on a time-continuous framework considered only a single
1084
1085 509 objective function, particularly the NSE or R2. The squared error statistics used for the NSE have been
1086
1087 510 shown to be biased toward high flows (Chilkoti et al., 2018). In addition, strong R2 values can still occur
1088
1089 511 even when systematic over- or under-prediction occurs (Krause et al., 2005). Beyond that, validation of
1090
1091 512 SWAT was only performed in a few studies as discussed in Section 3.1. The current testing approaches
1092
1093 513 that have been used for SWAT extreme assessments seem to be inconsistent. For instance, Yaduvanshi
1094
1095 514 et al. (2018) and Bacopoulos et al. (2017) validated SWAT’s capability to replicate specific flood events,
1096
1097 515 while Spellman et al. (2018) and Dakhlalla and Parajuli (2016) focused on the comparisons between
1098
516 simulated and measured peak flows. A new unified SWAT-based extreme assessment framework
1099
1100
517 should be developed to produce more reliable outputs that could also facilitate better comparison among
1101
1102
518 studies. Another future need is the incorporation of further calibration and validation for drought or
1103
1104
519 flood index calculations. For example, if the study objectives are related to drought assessment using
1105
1106 520 SSI, then calibration and validation should be done by comparing the SWAT-derived SSI values with
1107
1108 521 corresponding observations. Hence, the framework should consider both the general assessment and
1109
1110 522 extreme flows.
1111
1112 523
1113
1114
1115
1116
1117
1118
1119
1120 19
1121
1122
1123
1124
1125
524 4.2 SWAT low and high flows improvement
1126
1127
525 Tan et al. (2018) clearly showed that SWAT underestimated the volume of peak flow in the
1128
526 Kelantan River basin during the Big Yellow Flood event by 25.2% to 33.9%. Similarly, Bacopoulos et
1129
1130
527 al. (2017) found that SWAT poorly simulated the timing of peak flows in the lower St. Johns River
1131
1132
528 Basin in northeast Florida. Piniewski et al. (2017) reported SWAT underestimated low flows in small
1133
1134
529 sub-basins of the Vistula and Odra basins in Poland. These inaccurate representations of extreme flows,
1135
1136 530 along with similar results found in other studies (Zhang et al., 2015, Shrestha et al., 2019), indicate that
1137
1138 531 modification of SWAT’s internal algorithms are essential to enable the model to more accurately and
1139
1140 532 realistically simulate low and high flows.
1141
1142 533 The need for improving flood-plain deposition algorithms within SWAT has been noted in
1143
1144 534 previous research (Krysanova and Arnold, 2008, Yu et al., 2018a). As noted in Section 3.1, Yu et al.
1145
1146 535 (2018a) modified SWAT to more accurately simulate flood events using a sub-daily time step. Similar
1147
1148 536 modifications are needed to strengthen SWAT flood event predictions for daily time step simulations.
1149
1150 537 These issues have been considered during the development of SWAT+ (Bieger et al., 2017) However,
1151
1152 538 SWAT+ has not yet been widely applied in different regions across the world. Future research is needed
1153
1154 539 to compare the capability of SWAT+ versus previous SWAT versions, regarding replicating low and
1155
1156 540 peak flows for river basins representing different topographic, geographical and climatic conditions.
1157
1158 541 Another area of future research need is the development of an extreme flows module that
1159
1160 542 would be directly linked with the SWAT model. Such a module would feature functions capable of
1161
1162 543 providing extreme analyses that would be available in SWAT for users who want to study the impacts
1163
1164
544 of extreme flows. Some common extreme indices such as flow duration curve, 7Q10, SSI, 1-day
1165
545 maximum or minimum flow, and 7-day maximum or minimum flows could be considered as
1166
1167
546 components of an extreme flows module and should also be incorporated as part of the standard SWAT
1168
1169
547 outputs. Hydrological extremes classification can unfortunately be unclear. For example, Stewart et al.
1170
1171 548 (2015) defined extreme flows using a pre-defined threshold method where high flows are the months
1172
1173 549 with streamflow that are 100%, 125% and 150% greater than historical flows, whereas low flows are
1174
1175 550 25%, 50% and 75% lower than historical flows. By contrast, Tzoraki et al. (2013) applied the POT
1176
1177
1178
1179 20
1180
1181
1182
1183
551 method to classify floods into “usual”, “ecological” and “hazardous”. Therefore, future research is
1184
1185
552 needed to identify a simple and well-accepted extreme flows classification approach, which can be
1186
1187 553 plugged into a SWAT extreme flows module. A standardized extreme flows classification method
1188
1189 554 would facilitate comparison of flood or drought events in different regions.
1190
1191
555 4.3 SWAT with gridded precipitation data
1192
1193
556 Availability of satellite climate data offers another source of input data to SWAT modelling,
1194
1195
557 particularly for regions with limited ground-based climate observations. Modellers should test as many
1196
1197
558 available climate and other input data to choose the best data for a SWAT application (Abbaspour et
1198
1199 559 al., 2018). Reliable precipitation data is especially crucial as one of the main inputs to the SWAT model.
1200
1201 560 Observed precipitation data is almost always regarded as the most reliable precipitation input, but there
1202
1203 561 are many limitations in reality such as missing values, inhomogeneous measurements, uneven station
1204
1205 562 distribution and limited station coverage (Tan and Santo, 2018). For instance, Tan and Yang (2020)
1206
1207 563 reported that missing values of more than 20% in precipitation data would significantly impact on the
1208
1209 564 tropical streamflow simulation, especially in low-flow simulations. All of these issues might affect the
1210
1211 565 simulation of peak and low-flows.
1212
1213 566 To date, there are more than 30 global precipitation datasets available that can serve as
1214
1215 567 alternative precipitation data sources including gauge-based, reanalysis and satellite data (Sun et al.,
1216
1217 568 2018). These data sets are available for either long-term hydro-climatic extreme analysis (at least 30
1218
1219 569 years) or short-term events (up to 30-min temporal scale). Although the long-term National Centers for
1220
1221 570 Environmental (NCEP-CFSR) data are available at the SWAT website (SWAT, 2019a), it seems to be
1222
1223 571 less reliable in replicating rainfall for some regions and corresponding streamflow simulation by SWAT.
1224
1225 572 For example, the NCEP-CFSR data did not perform well in SWAT applications for river basins in China
1226
1227 573 (Yang et al., 2014), Brazil (Bressiani et al., 2015, Monteiro et al., 2016), Ethiopia (Roth and Lemann,
1228
574 2016) and Kenya and Tanzania (Alemayehu et al., 2017). In contrast, Tan et al. (2017) found that the
1229
1230
575 Asian Precipitation—Highly-Resolved Observational Data Integration towards Evaluation of Water
1231
1232
576 (APHRODITE) dataset (Yatagai et al., 2012, NCAR, 2019) performed better than the NCEP-CFSR for
1233
1234
577 two river basins simulated in SWAT in Malaysia. Yang et al. (2014) also report that APHRODITE
1235
1236
1237
1238 21
1239
1240
1241
1242
578 precipitation data resulted in improved streamflow predictions, relative to CFSR data, for a SWAT
1243
1244
579 application in central China. Therefore, more effort is required to test the reliability of freely available
1245
1246 580 precipitation products for SWAT extreme simulations. Moreover, integration of global, satellite and
1247
1248 581 observed precipitation data might produce better precipitation inputs for SWAT modelling.
1249
1250 582
1251
1252
583 4.4 SWAT climate projections
1253
1254
584 The CMIP3, CMIP5 and CMIP6 GCM climate projections provide a general view on how the
1255
1256
585 earth climate system may change in the future (IPCC, 2013). These projections have been widely
1257
1258 586 applied in SWAT for long-term hydro-climatic extreme assessments. Incorporation of the new CMIP6
1259
1260 587 GCM projections (LLNL, 2019) into SWAT applications will likely increase significantly in the near
1261
1262 588 future. As climate change analysis is one of the major applications of SWAT modelling (Tan et al.,
1263
1264 589 2019a, Gassman et al., 2014, CARD, 2019), providing the CMIP6 climate projections in a SWAT user
1265
1266 590 friendly format would facilitate increased usage of SWAT. A tool called Climate Model data for
1267
1268 591 hydrologic modelling (CMhyd) is designed to extract and bias correct climate projection data (Rathjens
1269
1270 592 et al., 2016) and is freely available at the SWAT website: https://swat.tamu.edu/software/cmhyd/. The
1271
1272 593 Climate Change Toolkit (CCT) has been developed for similar objectives (Ashraf Vaghefi et al., 2017)
1273
1274 594 and is comprised of five modules: (1) data download, (2) data extraction, (3) global climate data
1275
1276 595 management, (4) bias correction using statistical downscaling, (5) spatial interpolation of climate data,
1277
1278 596 and (6) Critical Consecutive Day Analyser (CCDA). However, these toolkits mainly focused on the
1279
1280 597 manipulation of CMIP5 GCMs. Hence, future research needs to be concentrated on the extraction and
1281
1282 598 bias correction of data from CMIP6 GCMs. Moreover, studies on the regional climate downscaling of
1283
1284 599 CMIP6 GCMs projections within SWAT are necessary to provide more accurate details for representing
1285
1286 600 localised extreme conditions.
1287
601 Climate projections are commonly known as one of the major uncertainties in any hydro-
1288
1289
602 climatic modelling focused on future conditions (Tan et al., 2014, Kundzewicz et al., 2018). These
1290
1291
603 uncertainties include the selection of future socio-economic scenarios, potential greenhouse gases and
1292
1293
604 aerosol emissions, GCMs climate sensitivity, regional climate models (RCMs), downscaling and bias
1294
1295
1296
1297 22
1298
1299
1300
1301
605 correction approaches. Climate projection uncertainty reduction or minimization is among the most
1302
1303
606 urgent future research needs (Kundzewicz et al., 2018), but obviously depends on the future decisions
1304
1305 607 on the degree of anthropogenic mitigation and adaptation. Therefore, a range of potential scenarios is
1306
1307 608 required to fully capture the possible future changes. However, the problem of mismatch of scales
1308
1309 609 between global-scale climate models and basin-scale hydrological models needs to studied and solved.
1310
1311 610 With the advent of greater computing power, higher resolution modelling studies are becoming
1312
1313 611 available and will reduce this problem but the computational intensity tends to mean that there are fewer
1314
1315 612 ensemble members of climate scenarios (Haarsma et al., 2016). Additionally, the use of RCMs can
1316
1317 613 overcome mismatch problems, but the usefulness of RCMs is generally limited to large-scale basins.
1318
1319 614 For small basins, only a limited subset of the RCMs grid points will overlay the entire basin, which will
1320
1321 615 not accurately represent the climate system for the entire basin. Thus, basin-scale climate models could
1322
1323 616 be among future climate change modeling development trends. Rapid development of knowledge,
1324
1325 617 computer and internet technologies could result in basin-scale climate models becoming easily available
1326
1327 618 in the future. Also, users are expecting large amounts of data to be freely available, and immediately
1328
1329 619 and easily accessible (Brodaric and Piasecki, 2016). Availability of bias-corrected multiple RCMs or
1330
1331 620 basin-scale climate data in SWAT format, that can be freely downloaded from the SWAT website,
1332
1333 621 would undoubtedly boost SWAT hydro-climatic extremes applications in the near future.
1334
622
1335
1336
1337 623 4.5 SWAT comparison with other models
1338
1339 624 Selection of the most accurate hydrological model could reduce the uncertainty for hydro-
1340
1341 625 climatic extreme simulations. However, the uncertainty of different hydrological models in regards to
1342
1343 626 estimating extreme streamflows is still poorly understood. Singh et al. (2005) compared the
1344
627 performance of SWAT with the Hydrological Simulation Program Fortran (HSPF) model in the
1345
1346
628 Iroquois River Watershed in Illinois and Indiana. They found that SWAT performed slightly better than
1347
1348
629 HSPF, particularly in low flows simulation. On the other hand, Chen et al. (2019b) concluded that both
1349
1350
630 SWAT and HSPF were unable to reproduce the extreme flows (i.e. annual maximum discharge and
1351
1352 631 annual 7-day minimum discharge) accurately in the Xitiaoxi River Basin that is located in eastern China.
1353
1354
1355
1356 23
1357
1358
1359
1360
632 Moreover, SWAT and HSPF resulted in different extreme trend directions, showing that a careful
1361
1362
633 consideration is needed when choosing hydrological models. Misuse of hydrological models when
1363
1364 634 simulating extreme streamflows might lead to erroneous decision making. Therefore, comparisons of
1365
1366 635 SWAT with other hydrological models for extreme flow simulations, including between SWAT2012
1367
1368 636 and SWAT+, should be conducted to identify the best model for further analysis for a respective
1369
1370 637 application.
1371
1372
1373 638 The application of artificial intelligence (sometime also known as machine learning or deep
1374
1375 639 learning) is increasing in many research fields. To date, a small set of studies have been conducted to
1376
1377 640 compare SWAT versus artificial intelligence (AI) models, which include comparisons for general flow
1378
1379 641 (Singh, 2016) and sediment yield (Singh et al., 2012) outputs. For extreme flow comparisons, Jimeno-
1380
1381 642 Sáez et al. (2018) compared the capability of the Artificial Neural Network (ANN) and SWAT models
1382
1383 643 in Spain. They found that the ANN model performed better for high flow simulations, whereas SWAT
1384
1385 644 was superior at simulating low flows. A similar finding was reported by Kim et al. (2015) for the
1386
1387 645 Taehwa River Watershed in the southern South Korea, where high flows were simulated more
1388
1389 646 accurately by both an ANN AI model and a Self Organizing Map (SOM) AI model, while SWAT was
1390
1391
647 better at capturing low flows. However, such comparisons are still relatively limited in the literature,
1392
648 and therefore require more investigation. Integration of AI within SWAT modelling is a research
1393
1394
649 direction that merits further exploration towards the goals of improving high flow simulations.
1395
1396
1397 650
1398
1399
1400 651 5.0 Summary
1401
1402 652 SWAT is now one of the top hydrological models that has been applied widely in ecological,
1403
1404 653 hydrological and environmental studies. Application of SWAT in hydro-climatic extreme studies has
1405
1406 654 increased rapidly in the past few years, particularly since 2017, as evidenced by the factor of three
1407
1408 655 increases in the number of publications compared to previous years. Nearly half of the reviewed studies
1409
1410 656 were conducted in the United States and China. 52% of the studies were conducted for river basins that
1411
1412
1413
1414
1415 24
1416
1417
1418
1419
657 drain more than 10,000 km2, showing the need to understand the effects of potential extreme changes
1420
1421
658 on the hydrology of medium to large scale basins.
1422
1423 659 Similar to other applications, calibration and validation of SWAT for extreme applications
1424
1425 660 were conducted mainly using the time-continuous approach by daily or monthly streamflow data. Only
1426
1427 661 a few studies further validated the model for specific flood events, peak flow, low flow, flow duration
1428
1429 662 curves and different precipitation intensity comparisons. Hydrological drought indices such as SSI,
1430
1431 663 SSMI, PDSI and 7Q10 are among the popular indices used in SWAT-based drought studies. SWAT
1432
1433 664 was used to study how land use changes such as water retention sites, hydropower development and
1434
1435 665 wetlands could reduce flood risk. Besides that, SWAT was also coupled with flood, economic and
1436
1437 666 habitat suitability models to further study the impacts of extreme events. Improvement of SWAT for
1438
1439 667 hydro-climatic extremes studies will require development of new drought or flood indices, calibration
1440
1441 668 method improvement, uncertainty analysis and SWAT modifications.
1442
1443 669 The lack of an extreme-based calibration and validation assessment is among the main
1444
1445 670 problems revealed in this study. A new unified SWAT-based extreme assessment framework that
1446
1447 671 combines both “traditional,” extreme flows and indices calculation should be developed for more
1448
1449 672 reliable analysis and comparison. As described by Bieger et al. (2017), several of these limitations have
1450
1451 673 been considered in the development of SWAT+, resulting in an improved flood plain module and
1452
674 streamflow-aquifer interaction. However, testing of SWAT+ has so far been limited to basins in the
1453
1454
675 United States (Bieger et al., 2017, Bieger et al., 2019, Wu et al., 2020) Further research is necessary to
1455
1456
676 compare the newly released SWAT+ to previous SWAT versions for extreme flow simulations, in
1457
1458
677 multiple basins in various regions that are characterized by different topographic, geographical and
1459
1460 678 climatic conditions.
1461
1462 679 Another key problem is that SWAT often does not accurately match peak and low flows, so
1463
1464 680 future research needs to focus on improving the replication of these extreme flows. Besides that,
1465
1466 681 development of an extreme flow module within an overall SWAT modelling system is another area of
1467
1468 682 future need. This would help address the need for reliable impact assessments in relatively short time-
1469
1470 683 frames that are required by policy makers. Also, a SWAT extreme module could help support more
1471
1472 684 consistent and improved comparisons between studies, which is important for understanding current
1473
1474 25
1475
1476
1477
1478
685 trends as stated in IPCC reports (IPCC, 2013, IPCC, 2012). Some common extreme indices such as
1479
1480
686 SSI and 7Q10, and a standardised extreme classification, should also be incorporated in the SWAT
1481
1482 687 output.
1483
1484 688 Availability of observations as inputs for SWAT modelling remains a big challenge in many
1485
1486 689 regions (Tan et al., 2019a, Bressiani et al., 2015, van Griensven et al., 2012). Application of global and
1487
1488 690 satellite precipitation data is becoming a new trend in SWAT modelling. Therefore, development of a
1489
1490 691 SWAT-based input data selection framework is essential to choose the most reliable data for extreme
1491
1492 692 simulations. Moreover, integration of global data, satellite products and ground-based observation is an
1493
1494 693 additional future need. Another future research direction is incorporation of CMIP6 GCMs in SWAT
1495
1496 694 hydro-climatic extreme studies. This would involve extraction, downscaling and bias correction
1497
1498 695 improvement, and availability on the SWAT website.
1499
1500 696
1501
1502
1503 697 Acknowledgement
1504
1505 698 This research was funded by the “IMpacts of PRecipitation from Extreme StormS – Malaysia
1506
1507 699 (IMPRESS – Malaysia)” Ministry of Higher Education Malaysia and NEWTON-NERC grant, Award
1508
1509 700 No. 203.PHUMANITI.6780001 (Malaysia) & NE/S002707/1 (UK); the Belt and Road Special
1510
1511 701 Foundation of the State Key Laboratory of Hydrology – Water Resources and Hydraulic Engineering
1512
1513 702 of China, Award No. 2018nkzd01; U.S. Department of Energy Initiative, Award No. DESC0016438,
1514
1515
703 “A Hierarchical Evaluation Framework for Assessing Climate Simulations Relevant to the Energy-
1516
704 Water-Land Nexus”, and by U.S. Department of Energy Initiative, Award No. DESC0016605, “An
1517
1518
705 Integrated Assessment of Regional Climate-Water-Energy-Land-Decision Modeling.”.
1519
1520
706
1521
1522
1523
707 References
1524
1525
708 Abbaspour, K. C., Vaghefi, S. A. & Srinivasan, R. 2018. A Guideline for Successful Calibration and
1526
1527
709 Uncertainty Analysis for Soil and Water Assessment: A Review of Papers from the 2016
1528
1529 710 International SWAT Conference. Water, 10, 18.
1530
1531
1532
1533 26
1534
1535
1536
1537
711 Ahn, K.-H. & Merwade, V. 2017. The effect of land cover change on duration and severity of high and
1538
1539
712 low flows. Hydrological Processes, 31, 133-149.
1540
1541 713 Ahn, S., Abudu, S., Sheng, Z. & Mirchi, A. 2018. Hydrologic impacts of drought-adaptive agricultural
1542
1543 714 water management in a semi-arid river basin: Case of Rincon Valley, New Mexico.
1544
1545 715 Agricultural Water Management, 209, 206-218.
1546
1547 716 Ahn, S. R., Jeong, J. H. & Kim, S. J. 2016. Assessing drought threats to agricultural water supplies
1548
1549 717 under climate change by combining the SWAT and MODSIM models for the Geum River basin,
1550
1551 718 South Korea. Hydrological Sciences Journal, 61, 2740-2753.
1552
1553 719 Alemayehu, T., Kilonzo, F., Van Griensven, A. & Bauwens, W. 2017. Evaluation and application of
1554
1555 720 alternative rainfall data sources for forcing hydrologic models in the Mara Basin. Hydrology
1556
1557 721 Research, 49, 1271-1282.
1558
1559 722 Alodah, A. & Seidou, O. 2019. Assessment of Climate Change Impacts on Extreme High and Low
1560
1561 723 Flows: An Improved Bottom-Up Approach. Water, 11, 1236.
1562
1563 724 Analytics, C. 2019. Web of Science platform: Web of Science Core Collection. Clarivate Analytics:
1564
1565 725 Philadelphia, PA. Available at: https://clarivate.libguides.com/webofscienceplatform/woscc.
1566
1567 726 [Online]. [Accessed].
1568
1569 727 Angelidis, P., Kotsikas, M. & Kotsovinos, N. 2010. Management of Upstream Dams and Flood
1570
728 Protection of the Transboundary River Evros/Maritza. Water Resources Management, 24,
1571
1572
729 2467-2484.
1573
1574
730 Arias, M. E., Piman, T., Lauri, H., Cochrane, T. A. & Kummu, M. 2014. Dams on Mekong tributaries
1575
1576
731 as significant contributors of hydrological alterations to the Tonle Sap Floodplain in Cambodia.
1577
1578 732 Hydrol. Earth Syst. Sci., 18, 5303-5315.
1579
1580 733 Arnold, J., Moriasi, D., Gassman, P., C. Abbaspour, K., White, M., Srinivasan, R., Santhi, C., D. Harmel,
1581
1582 734 R., Van Griensven, A., Van Liew, M., Kannan, N. & Jha, M. 2012a. SWAT: Model use,
1583
1584 735 calibration, and validation. Transactions of the ASABE, 55, 1491-1508.
1585
1586 736 Arnold, J. G., Kiniry, J. R., Srinivasan, R., Williams, J. R., Haney, E. B. & Neitsch, S. L. 2012b. Soil
1587
1588 737 and Water Assessment Tool Input/Output File Documentation: Version 2012 (Texas Water
1589
1590
1591
1592 27
1593
1594
1595
1596
738 Resources Institute TR-439), Temple, Texas: USDA-ARS, Grassland, Soil and Water Research
1597
1598
739 Laboratory, and Texas AgriLife Research, Blackland Research and Extension Center.
1599
1600 740 Arnold, J. G., Srinivasan, R., Muttiah, R. S. & Williams, J. R. 1998. Large area hydrologic modeling
1601
1602 741 and assessment part I: Model development. JAWRA Journal of the American Water Resources
1603
1604 742 Association, 34, 73-89.
1605
1606 743 Arunyanart, N., Limsiri, C. & Uchaipichat, A. 2017. Flood Hazards in the Chi River Basin, Thailand:
1607
1608 744 Impact Management of Climate Change. Applied Ecology and Environmental Research, 15,
1609
1610 745 841-861.
1611
1612 746 Ashraf Vaghefi, S., Abbaspour, N., Kamali, B. & Abbaspour, K. C. 2017. A toolkit for climate change
1613
1614 747 analysis and pattern recognition for extreme weather conditions – Case study: California-Baja
1615
1616 748 California Peninsula. Environmental Modelling & Software, 96, 181-198.
1617
1618 749 Ashraf Vaghefi, S., Mousavi, S. J., Abbaspour, K. C., Srinivasan, R. & Yang, H. 2014. Analyses of the
1619
1620 750 impact of climate change on water resources components, drought and wheat yield in semiarid
1621
1622 751 regions: Karkheh River Basin in Iran. Hydrological Processes, 28, 2018-2032.
1623
1624 752 Bacopoulos, P., Tang, Y., Wang, D. & Hagen, S. C. 2017. Integrated Hydrologic-Hydrodynamic
1625
1626 753 Modeling of Estuarine-Riverine Flooding: 2008 Tropical Storm Fay. Journal of Hydrologic
1627
1628 754 Engineering, 22, 04017022.
1629
755 Bayissa, Y., Maskey, S., Tadesse, T., Van Andel, S. J., Moges, S., Van Griensven, A. & Solomatine,
1630
1631
756 D. 2018. Comparison of the Performance of Six Drought Indices in Characterizing Historical
1632
1633
757 Drought for the Upper Blue Nile Basin, Ethiopia. Geosciences, 8, 81.
1634
1635
758 Bieger, K., Arnold, J. G., Rathjens, H., White, M. J., Bosch, D. D. & Allen, P. M. 2019. Representing
1636
1637 759 the Connectivity of Upland Areas to Floodplains and Streams in SWAT+. JAWRA Journal of
1638
1639 760 the American Water Resources Association, 55, 578-590.
1640
1641 761 Bieger, K., Arnold, J. G., Rathjens, H., White, M. J., Bosch, D. D., Allen, P. M., Volk, M. & Srinivasan,
1642
1643 762 R. 2017. Introduction to SWAT+, A Completely Restructured Version of the Soil and Water
1644
1645 763 Assessment Tool. JAWRA Journal of the American Water Resources Association, 53, 115-130.
1646
1647
1648
1649
1650
1651 28
1652
1653
1654
1655
764 Boithias, L., Sauvage, S., Lenica, A., Roux, H., Abbaspour, K. C., Larnier, K., Dartus, D. & Sanchez-
1656
1657
765 Perez, J. M. 2017. Simulating Flash Floods at Hourly Time-Step Using the SWAT Model.
1658
1659 766 Water, 9, 25.
1660
1661 767 Borah, D. & Bera, M. 2004. Watershed-scale hydrologic and nonpoint-source pollution models: Review
1662
1663 768 of applications. Transactions of the ASAE, 47, 789-803.
1664
1665 769 Bressiani, D. A., Gassman, P. W., Fernandes, J. G., Garbossa, L. H. P., Srinivasan, R., Bonumá, N. B.
1666
1667 770 & Mendiondo, E. M. 2015. Review of Soil and Water Assessment Tool (SWAT) applications
1668
1669 771 in Brazil: Challenges and prospects. International Journal of Agricultural and Biological
1670
1671 772 Engineering, 8, 9-35.
1672
1673 773 Brighenti, T. M., Bonumá, N. B., Srinivasan, R. & Chaffe, P. L. B. 2019. Simulating sub-daily
1674
1675 774 hydrological process with SWAT: a review. Hydrological Sciences Journal, 64, 1415-1423.
1676
1677 775 Brodaric, B. & Piasecki, M. 2016. Editorial: Water data networks: Foundations, technologies and
1678
1679 776 systems, implementations, and uses. Journal of Hydroinformatics, 18, 149-151.
1680
1681 777 Brown, S. C., Versace, V. L., Lester, R. E. & Todd Walter, M. 2015. Assessing the impact of drought
1682
1683 778 and forestry on streamflows in south-eastern Australia using a physically based hydrological
1684
1685 779 model. Environmental Earth Sciences, 74, 6047-6063.
1686
1687 780 Cai, X., Zeng, R., Kang, W. H., Song, J. & Valocchi, A. J. 2015. Strategic Planning for Drought
1688
781 Mitigation under Climate Change. Journal of Water Resources Planning and Management, 141,
1689
1690
782 04015004.
1691
1692
783 Campbell, A., Pradhanang, S. M., Kouhi Anbaran, S., Sargent, J., Palmer, Z. & Audette, M. 2018.
1693
1694
784 Assessing the impact of urbanization on flood risk and severity for the Pawtuxet watershed,
1695
1696 785 Rhode Island. Lake and Reservoir Management, 34, 74-87.
1697
1698 786 Card 2019. SWAT Literature Database for Peer-Reviewed Journal Articles; Center for Agricultural and
1699
1700 787 Rural Development: Iowa State University, Ames, IA, USA
1701
1702 788 Chattopadhyay, S., Edwards, D. R., Yu, Y. & Hamidisepehr, A. 2017. An Assessment of Climate
1703
1704 789 Change Impacts on Future Water Availability and Droughts in the Kentucky River Basin.
1705
1706 790 Environmental Processes-an International Journal, 4, 477-507.
1707
1708
1709
1710 29
1711
1712
1713
1714
791 Chen, F. & Li, J. 2016. Quantifying drought and water scarcity: a case study in the Luanhe river basin.
1715
1716
792 Natural Hazards, 81, 1913-1927.
1717
1718 793 Chen, Q., Chen, H., Wang, J., Zhao, Y., Chen, J. & Xu, C. 2019a. Impacts of Climate Change and Land-
1719
1720 794 Use Change on Hydrological Extremes in the Jinsha River Basin. Water, 11, 1398.
1721
1722 795 Chen, Y., Xu, C.-Y., Chen, X., Xu, Y., Yin, Y., Gao, L. & Liu, M. 2019b. Uncertainty in simulation of
1723
1724 796 land-use change impacts on catchment runoff with multi-timescales based on the comparison
1725
1726 797 of the HSPF and SWAT models. Journal of Hydrology, 573, 486-500.
1727
1728 798 Cheng, C., Yang, Y. C. E., Ryan, R., Yu, Q. & Brabec, E. 2017. Assessing climate change-induced
1729
1730 799 flooding mitigation for adaptation in Boston’s Charles River watershed, USA. Landscape and
1731
1732 800 Urban Planning, 167, 25-36.
1733
1734 801 Chilkoti, V., Bolisetti, T. & Balachandar, R. 2018. Multi-objective autocalibration of SWAT model for
1735
1736 802 improved low flow performance for a small snowfed catchment. Hydrological Sciences Journal,
1737
1738 803 63, 1482-1501.
1739
1740 804 Chinnasamy, P., Muthuwatta, L., Eriyagama, N., Pavelic, P. & Lagudu, S. 2018. Modeling the potential
1741
1742 805 for floodwater recharge to offset groundwater depletion: a case study from the Ramganga basin,
1743
1744 806 India. Sustainable Water Resources Management, 4, 331-344.
1745
1746 807 Chiogna, G., Marcolini, G., Liu, W., Pérez Ciria, T. & Tuo, Y. 2018. Coupling hydrological modeling
1747
808 and support vector regression to model hydropeaking in alpine catchments. Science of The Total
1748
1749
809 Environment, 633, 220-229.
1750
1751
810 Chou, C., Chiang, J. C. H., Lan, C. W., Chung, C. H., Liao, Y. C. & Lee, C. J. 2013. Increase in the
1752
1753
811 range between wet and dry season precipitation. Nature Geoscience, 6, 263-267.
1754
1755 812 Cloke, H. L. & Pappenberger, F. 2009. Ensemble flood forecasting: A review. Journal of Hydrology,
1756
1757 813 375, 613-626.
1758
1759 814 Coffey, M. E., Workman, S. R., Taraba, J. L. & W. Fogle, A. W. 2004. Statistical Procedures for
1760
1761 815 Evaluating Daily and Monthly Hydrologic Model Predictions. Transactions of the ASAE, 47,
1762
1763 816 59-68.
1764
1765
1766
1767
1768
1769 30
1770
1771
1772
1773
817 Cohen Liechti, T., Matos, J. P., Ferràs Segura, D., Boillat, J.-L. & Schleiss, A. J. 2014. Hydrological
1774
1775
818 modelling of the Zambezi River Basin taking into account floodplain behaviour by a modified
1776
1777 819 reservoir approach. International Journal of River Basin Management, 12, 29-41.
1778
1779 820 D. White, E., W. Feyereisen, G., L. Veith, T. & D. Bosch, D. 2009. Improving Daily Water Yield
1780
1781 821 Estimates in the Little River Watershed: SWAT Adjustments. Transactions of the ASABE, 52,
1782
1783 822 69-79.
1784
1785 823 Dakhlalla, A. O. & Parajuli, P. B. 2016. Evaluation of the Best Management Practices at the Watershed
1786
1787 824 Scale to Attenuate Peak Streamflow Under Climate Change Scenarios. Water Resources
1788
1789 825 Management, 30, 963-982.
1790
1791 826 Daniel, E. B. 2011. Watershed Modeling and its Applications: A State-of-the-Art Review. The Open
1792
1793 827 Hydrology Journal, 5, 26-50.
1794
1795 828 Dash, S. S., Sahoo, B. & Raghuwanshi, N. S. 2019. A SWAT-Copula based approach for monitoring
1796
1797 829 and assessment of drought propagation in an irrigation command. Ecological Engineering, 127,
1798
1799 830 417-430.
1800
1801 831 Deltares 2019. SOBEK Hydrodynamics, Rainfall Runoff and Real Time Control User Manual. Deltares,
1802
1803 832 Boussinesqweg 1, 2629 HV Delft, P.O. 177, 2600 MH Delft, the Netherlands.
1804
1805 833 Dhi 2017. MIKE Flood User Manual. Agern Allé 5, DK-2970 Hørsholm, Denmark.
1806
834 Dile, Y. T., Daggupati, P., George, C., Srinivasan, R. & Arnold, J. 2016. Introducing a new open source
1807
1808
835 GIS user interface for the SWAT model. Environmental Modelling & Software, 85, 129-138.
1809
1810
836 Douglas-Mankin, K. R., Srinivasan, R. & Arnold, A. J. 2010. Soil and Water Assessment Tool (SWAT)
1811
1812
837 Model: Current Developments and Applications. Transactions of the ASABE, 53, 1423-1431.
1813
1814 838 Duan, Y., Liu, T., Meng, F., Luo, M., Frankl, A., De Maeyer, P., Bao, A., Kurban, A. & Feng, X. 2018.
1815
1816 839 Inclusion of Modified Snow Melting and Flood Processes in the SWAT Model. Water, 10, 1715.
1817
1818 840 Engel, B., Storm, D., White, M., Arnold, J. & Arabi, M. 2007. A Hydrologic⁄Water Quality Model
1819
1820 841 Application Protocol. JAWRA Journal of the American Water Resources Association, 43, 1223-
1821
1822 842 1236.
1823
1824
1825
1826
1827
1828 31
1829
1830
1831
1832
843 Esfahanian, E., Nejadhashemi, A. P., Abouali, M., Adhikari, U., Zhang, Z., Daneshvar, F. & Herman,
1833
1834
844 M. R. 2017. Development and evaluation of a comprehensive drought index. Journal of
1835
1836 845 Environmental Management, 185, 31-43.
1837
1838 846 Esfahanian, E., Nejadhashemi, A. P., Abouali, M., Daneshvar, F., Alireza, A. R., Herman, M. R. &
1839
1840 847 Tang, Y. 2016. Defining drought in the context of stream health. Ecological Engineering, 94,
1841
1842 848 668-681.
1843
1844 849 Francesconi, W., Srinivasan, R., Perez-Minana, E., Willcock, S. P. & Quintero, M. 2016. Using the Soil
1845
1846 850 and Water Assessment Tool (SWAT) to model ecosystem services: A systematic review.
1847
1848 851 Journal of Hydrology, 535, 625-636.
1849
1850 852 Fu, B., Merritt, W. S., Croke, B. F. W., Weber, T. R. & Jakeman, A. J. 2019. A review of catchment-
1851
1852 853 scale water quality and erosion models and a synthesis of future prospects. Environmental
1853
1854 854 Modelling & Software, 114, 75-97.
1855
1856 855 Fu, G., Barber, M. E. & Chen, S. 2010. Hydro-climatic variability and trends in Washington State for
1857
1858 856 the last 50 years. Hydrological Processes, 24, 866-878.
1859
1860 857 Gao, Y. Q., Guo, Z. C., Wang, D. D., Zhang, Z. X. & Liu, Y. P. 2018. Multivariate Flood Risk Analysis
1861
1862 858 at a Watershed Scale Considering Climatic Factors. Water, 10, 15.
1863
1864 859 Gassman, P. W., Reyes, M. R., Green, C. H. & Arnold, J. G. 2007. The Soil and Water Assessment
1865
860 Tool: Historical Development, Applications, and Future Research Directions. Transactions of
1866
1867
861 the ASABE, 50, 1211-1250.
1868
1869
862 Gassman, P. W., Sadeghi, A. M. & Srinivasan, R. 2014. Applications of the SWAT Model Special
1870
1871
863 Section: Overview and Insights. Journal of Environmental Quality, 43, 1-8.
1872
1873 864 Gassman, P. W. & Wang, Y. 2015. IJABE SWAT Special Issue: Innovative modeling solutions for
1874
1875 865 water resource problems. International Journal of Agricultural and Biological Engineering, 8,
1876
1877 866 1-8.
1878
1879 867 Gharib, A., Davies, E. G. R., Goss, G. G. & Faramarzi, M. 2017. Assessment of the Combined Effects
1880
1881 868 of Threshold Selection and Parameter Estimation of Generalized Pareto Distribution with
1882
1883 869 Applications to Flood Frequency Analysis. Water, 9, 692.
1884
1885
1886
1887 32
1888
1889
1890
1891
870 Gies, L., Agusdinata, D. B. & Merwade, V. 2014. Drought adaptation policy development and
1892
1893
871 assessment in East Africa using hydrologic and system dynamics modeling. Natural Hazards,
1894
1895 872 74, 789-813.
1896
1897 873 Giorgi, F., Coppola, E. & Raffaele, F. 2018. Threatening levels of cumulative stress due to
1898
1899 874 hydroclimatic extremes in the 21st century. npj Climate and Atmospheric Science, 1, 18.
1900
1901 875 Giorgi, F., Jones, C. & Asrar, G. 2009. Addressing climate information needs at the regional level: the
1902
1903 876 CORDEX framework. WMO Bulletin of the American Meteorological Society, 58, 175-183.
1904
1905 877 Green, N. H. & Ampt, C. A. 1911. Flow of air and water through soils. Journal of Agricultural Science,
1906
1907 878 4, 1-24.
1908
1909 879 Haarsma, R. J., Roberts, M. J., Vidale, P. L., Senior, C. A., Bellucci, A., Bao, Q., Chang, P., Corti, S.,
1910
1911 880 Fučkar, N. S., Guemas, V., Von Hardenberg, J., Hazeleger, W., Kodama, C., Koenigk, T.,
1912
1913 881 Leung, L. R., Lu, J., Luo, J. J., Mao, J., Mizielinski, M. S., Mizuta, R., Nobre, P., Satoh, M.,
1914
1915 882 Scoccimarro, E., Semmler, T., Small, J. & Von Storch, J. S. 2016. High Resolution Model
1916
1917 883 Intercomparison Project (HighResMIP v1.0) for CMIP6. Geosci. Model Dev., 9, 4185-4208.
1918
1919 884 Harmel, R. D., Baffaut, C. & Douglas-Mankin, K. R. 2018. Review and development of ASABE
1920
1921 885 Engineering Practice 621: “Guidelines for calibrating, validating, and evaluating hydrologic
1922
1923 886 and water quality models”. Transactions of the ASABE, 61, 1393-1401.
1924
887 Hou, A. Y., Kakar, R. K., Neeck, S., Azarbarzin, A. A., Kummerow, C. D., Kojima, M., Oki, R.,
1925
1926
888 Nakamura, K. & Iguchi, T. 2014. The Global Precipitation Measurement Mission. Bulletin of
1927
1928
889 the American Meteorological Society, 95, 701-722.
1929
1930
890 Hoyos, N., Correa-Metrio, A., Jepsen, S. M., Wemple, B., Valencia, S., Marsik, M., Doria, R., Escobar,
1931
1932 891 J., Restrepo, J. C. & Velez, M. I. 2019. Modeling Streamflow Response to Persistent Drought
1933
1934 892 in a Coastal Tropical Mountainous Watershed, Sierra Nevada De Santa Marta, Colombia.
1935
1936 893 Water, 11, 94.
1937
1938 894 Huang, X. D., Wang, L., Han, P. P. & Wang, W. C. 2018. Spatial and Temporal Patterns in
1939
1940 895 Nonstationary Flood Frequency across a Forest Watershed: Linkage with Rainfall and Land
1941
1942 896 Use Types. Forests, 9, 20.
1943
1944
1945
1946 33
1947
1948
1949
1950
897 Huffman, G. J., Bolvin, D. T., Nelkin, E. J., Wolff, D. B., Adler, R. F., Gu, G., Hong, Y., Bowman, K.
1951
1952
898 P. & Stocker, E. F. 2007. The TRMM Multisatellite Precipitation Analysis (TMPA): Quasi-
1953
1954 899 Global, Multiyear, Combined-Sensor Precipitation Estimates at Fine Scales. Journal of
1955
1956 900 Hydrometeorology, 8, 38-55.
1957
1958 901 Hwang, T. H., Kim, B. S., Kim, H. S. & Seoh, B. H. 2006. The estimation of soil moisture index by
1959
1960 902 SWAT model and drought monitoring. Journal of the Korean Society of Civil Engineers, 26,
1961
1962 903 345-354.
1963
1964 904 Igarashi, K., Koichiro, K., Tanaka, N. & Aranyabhaga, N. 2019. Prediction of the Impact of Climate
1965
1966 905 Change and Land Use Change on Flood Discharge in the Song Khwae District, Nan Province,
1967
1968 906 Thailand. Journal of Climate Change, 5, 1-8.
1969
1970 907 Ipcc 2012. Managing the Risks of Extreme Events and Disasters to Advance Climate Change
1971
1972 908 Adaptation. A Special Report of Working Groups I and II of the Intergovernmental Panel on
1973
1974 909 Climate Change In: FIELD, C. B., V. BARROS, T.F. STOCKER, D. QIN, D.J. DOKKEN,
1975
1976 910 K.L. EBI, M.D. MASTRANDREA, K.J. MACH, G.-K. PLATTNER, S.K. ALLEN, M.
1977
1978 911 TIGNOR, AND P.M. MIDGLEY (EDS.) (ed.). Cambridge University Press, Cambridge,
1979
1980 912 United Kingdom and New York, NY, USA.
1981
1982 913 Ipcc 2013. Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to the
1983
914 Fifth Assessment Report of the Intergovernmental Panel on Climate Change Cambridge
1984
1985
915 University Press, Cambridge, United Kingdom and New York, NY, USA, .
1986
1987
916 Iqbal, M. S., Dahri, Z. H., Querner, E. P., Khan, A. & Hofstra, N. 2018. Impact of Climate Change on
1988
1989
917 Flood Frequency and Intensity in the Kabul River Basin. Geosciences, 8, 114.
1990
1991 918 Jain, V. K., Pandey, R. P. & Jain, M. K. 2015. Spatio-temporal assessment of vulnerability to drought.
1992
1993 919 Natural Hazards, 76, 443-469.
1994
1995 920 Jamrussri, S. & Toda, Y. 2017. Simulating past severe flood events to evaluate the effectiveness of
1996
1997 921 nonstructural flood countermeasures in the upper Chao Phraya River Basin, Thailand. Journal
1998
1999 922 of Hydrology: Regional Studies, 10, 82-94.
2000
2001
2002
2003
2004
2005 34
2006
2007
2008
2009
923 Javaheri, A. & Babbar-Sebens, M. 2014. On comparison of peak flow reductions, flood inundation
2010
2011
924 maps, and velocity maps in evaluating effects of restored wetlands on channel flooding.
2012
2013 925 Ecological Engineering, 73, 132-145.
2014
2015 926 Jeong, J., Kannan, N., Arnold, J., Glick, R., Gosselink, L. & Srinivasan, R. 2010. Development and
2016
2017 927 Integration of Sub-hourly Rainfall–Runoff Modeling Capability Within a Watershed Model.
2018
2019 928 Water Resources Management, 24, 4505-4527.
2020
2021 929 Jimeno-Sáez, P., Senent-Aparicio, J., Pérez-Sánchez, J. & Pulido-Velazquez, D. 2018. A Comparison
2022
2023 930 of SWAT and ANN Models for Daily Runoff Simulation in Different Climatic Zones of
2024
2025 931 Peninsular Spain. Water, 10, 192.
2026
2027 932 Jodar-Abellan, A., Valdes-Abellan, J., Pla, C. & Gomariz-Castillo, F. 2019. Impact of land use changes
2028
2029 933 on flash flood prediction using a sub-daily SWAT model in five Mediterranean ungauged
2030
2031 934 watersheds (SE Spain). Science of The Total Environment, 657, 1578-1591.
2032
2033 935 Kamali, B., Houshmand Kouchi, D., Yang, H. & Abbaspour, K. C. 2017. Multilevel Drought Hazard
2034
2035 936 Assessment under Climate Change Scenarios in Semi-Arid Regions—A Case Study of the
2036
2037 937 Karkheh River Basin in Iran. Water, 9, 241.
2038
2039 938 Kang, H. & Sridhar, V. 2017a. Combined statistical and spatially distributed hydrological model for
2040
2041 939 evaluating future drought indices in Virginia. Journal of Hydrology: Regional Studies, 12, 253-
2042
940 272.
2043
2044
941 Kang, H. & Sridhar, V. 2017b. Description of future drought indices in Virginia. Data in Brief, 14, 278-
2045
2046
942 290.
2047
2048
943 Kang, H. & Sridhar, V. 2018a. Assessment of Future Drought Conditions in the Chesapeake Bay
2049
2050 944 Watershed. JAWRA Journal of the American Water Resources Association, 54, 160-183.
2051
2052 945 Kang, H. & Sridhar, V. 2018b. Improved Drought Prediction Using Near Real-Time Climate Forecasts
2053
2054 946 and Simulated Hydrologic Conditions. Sustainability, 10, 1799.
2055
2056 947 Kehew, A. E., Milewski, A. & Soliman, F. 2010. Reconstructing an extreme flood from boulder
2057
2058 948 transport and rainfall–runoff modelling: Wadi Isla, South Sinai, Egypt. Global and Planetary
2059
2060 949 Change, 70, 64-75.
2061
2062
2063
2064 35
2065
2066
2067
2068
950 Kharel, G. & Kirilenko, A. 2015. Considering Climate Change in the Estimation of Long-Term Flood
2069
2070
951 Risks of Devils Lake in North Dakota. JAWRA Journal of the American Water Resources
2071
2072 952 Association, 51, 1221-1234.
2073
2074 953 Kharel, G. & Kirilenko, A. 2018. Comparing CMIP-3 and CMIP-5 climate projections on flooding
2075
2076 954 estimation of Devils Lake of North Dakota, USA. PeerJ, 6, e4711.
2077
2078 955 Kharel, G., Zheng, H. & Kirilenko, A. 2016. Can land-use change mitigate long-term flood risks in the
2079
2080 956 Prairie Pothole Region? The case of Devils Lake, North Dakota, USA. Regional Environmental
2081
2082 957 Change, 16, 2443-2456.
2083
2084 958 Kim, M., Baek, S., Ligaray, M., Pyo, J., Park, M. & Cho, K. H. 2015. Comparative Studies of Different
2085
2086 959 Imputation Methods for Recovering Streamflow Observation. Water, 7, 6847-6860.
2087
2088 960 Kim, N. W. & Lee, J. 2008. Temporally weighted average curve number method for daily runoff
2089
2090 961 simulation. Hydrological Processes, 22, 4936-4948.
2091
2092 962 Kim, S., Parhi, P., Jun, H. & Lee, J. 2018. Evaluation of Drought Severity with a Bayesian Network
2093
2094 963 Analysis of Multiple Drought Indices. Journal of Water Resources Planning and Management,
2095
2096 964 144, 05017016.
2097
2098 965 Krause, P., Boyle, D. P. & Bäse, F. 2005. Comparison of different efficiency criteria for hydrological
2099
2100 966 model assessment. Adv. Geosci., 5, 89-97.
2101
967 Krysanova, V. & Arnold, J. G. 2008. Advances in ecohydrological modelling with SWAT-a review.
2102
2103
968 Hydrological Sciences Journal-Journal Des Sciences Hydrologiques, 53, 939-947.
2104
2105
969 Krysanova, V. & White, M. 2015. Advances in water resources assessment with SWAT—an overview.
2106
2107
970 Hydrological Sciences Journal, 60, 771-783.
2108
2109 971 Kumar, B. & Lakshmi, V. 2018. Accessing the capability of TRMM 3B42 V7 to simulate streamflow
2110
2111 972 during extreme rain events: Case study for a Himalayan River Basin. Journal of Earth System
2112
2113 973 Science, 127, 15.
2114
2115 974 Kundzewicz, Z. W., Krysanova, V., Benestad, R. E., Hov, Ø., Piniewski, M. & Otto, I. M. 2018.
2116
2117 975 Uncertainty in climate change impacts on water resources. Environmental Science & Policy,
2118
2119 976 79, 1-8.
2120
2121
2122
2123 36
2124
2125
2126
2127
977 Kuntiyawichai, K., Schultz, B., Uhlenbrook, S., Suryadi, F. X. & Van Griensven, A. 2011. Comparison
2128
2129
978 of flood management options for the Yang River Basin, Thailand. Irrigation and Drainage, 60,
2130
2131 979 526-543.
2132
2133 980 Lee, J.-H., Park, S.-Y., Kim, J.-S., Sur, C. & Chen, J. 2018. Extreme drought hotspot analysis for
2134
2135 981 adaptation to a changing climate: Assessment of applicability to the five major river basins of
2136
2137 982 the Korean Peninsula. International Journal of Climatology, 38, 4025-4032.
2138
2139 983 Lee, J. E., Heo, J.-H., Lee, J. & Kim, N. W. 2017. Assessment of Flood Frequency Alteration by Dam
2140
2141 984 Construction via SWAT Simulation. Water, 9, 264.
2142
2143 985 Leon, A. S., Kanashiro, E. A., Valverde, R. & Sridhar, V. 2014. Dynamic Framework for Intelligent
2144
2145 986 Control of River Flooding: Case Study. Journal of Water Resources Planning and Management,
2146
2147 987 140, 258-268.
2148
2149 988 Leta, O. T. & Bauwens, W. 2018. Assessment of the Impact of Climate Change on Daily Extreme Peak
2150
2151 989 and Low Flows of Zenne Basin in Belgium. Hydrology, 5, 38.
2152
2153 990 Leta, O. T., El-Kadi, A. I. & Dulai, H. 2018. Impact of Climate Change on Daily Streamflow and Its
2154
2155 991 Extreme Values in Pacific Island Watersheds. Sustainability, 10, 2057.
2156
2157 992 Li, D. C., Qu, S. M., Shi, P., Chen, X. Q., Xue, F., Gou, J. F. & Zhang, W. H. 2018a. Development and
2158
2159 993 Integration of Sub-Daily Flood Modelling Capability within the SWAT Model and a
2160
994 Comparison with XAJ Model. Water, 10, 17.
2161
2162
995 Li, J. Z., Guo, Y. G., Wang, Y. X., Lu, S. L. & Chen, X. 2018b. Drought Propagation Patterns under
2163
2164
996 Naturalized Condition Using Daily Hydrometeorological Data. Advances in Meteorology, 14.
2165
2166
997 Li, P., Omani, N., Chaubey, I. & Wei, X. 2017. Evaluation of Drought Implications on Ecosystem
2167
2168 998 Services: Freshwater Provisioning and Food Provisioning in the Upper Mississippi River Basin.
2169
2170 999 International Journal of Environmental Research and Public Health, 14, 496.
2171
2172 1000 Llnl 2019. CMIP6 - Coupled Model Intercomparison Project Phase 6. Lawrence Livermore National
2173
2174 1001 Laboratory, Livermore, CA. Available: https://pcmdi.llnl.gov/CMIP6/. [Accessed 17 October
2175
2176 1002 2019].
2177
2178 1003 Lopes, C. L., Alves, F. L. & Dias, J. M. 2017. Flood risk assessment in a coastal lagoon under present
2179
2180 1004 and future scenarios: Ria de Aveiro case study. Natural Hazards, 89, 1307-1325.
2181
2182 37
2183
2184
2185
2186
1005 Lu, J., Cui, X., Chen, X., Sauvage, S. & Sanchez Perez, J.-M. 2017. Evaluation of hydrological response
2187
2188
1006 to extreme climate variability using SWAT model: application to the Fuhe basin of Poyang
2189
2190 1007 Lake watershed, China. Hydrology Research, 48, 1730-1744.
2191
2192 1008 Lweendo, M. K., Lu, B., Wang, M., Zhang, H. & Xu, W. 2017. Characterization of Droughts in Humid
2193
2194 1009 Subtropical Region, Upper Kafue River Basin (Southern Africa). Water, 9, 242.
2195
2196 1010 Maghsood, F. F., Moradi, H., Bavani, A. R. M., Panahi, M., Berndtsson, R. & Hashemi, H. 2019.
2197
2198 1011 Climate Change Impact on Flood Frequency and Source Area in Northern Iran under CMIP5
2199
2200 1012 Scenarios. Water, 11, 21.
2201
2202 1013 Mcdaniel, R. L., Munster, C. & Cothren, J. T. 2017a. Crop and Location Specific Agricultural Drought
2203
2204 1014 Quantification: Part I. Method Development. Transactions of the ASABE, 60, 721-728.
2205
2206 1015 Mcdaniel, R. L., Munster, C. & Cothren, J. T. 2017b. Crop and Location Specific Agricultural Drought
2207
2208 1016 Quantification: Part II. Case Study. Transactions of the ASABE, 60, 729-739.
2209
2210 1017 Mcdaniel, R. L., Munster, C. & Nielsen-Gammon, J. 2017c. Crop and Location Specific Agricultural
2211
2212 1018 Drought Quantification: Part III. Forecasting Water Stress and Yield Trends. Transactions of
2213
2214 1019 the ASABE, 60, 741-752.
2215
2216 1020 Mearns, L. O., Gutowski, W., Jones, R., Leung, R., Mcginnis, S., Nunes, A. & Qian, Y. 2009. A
2217
2218 1021 Regional Climate Change Assessment Program for North America. Eos, Transactions
2219
1022 American Geophysical Union, 90, 311-311.
2220
2221
1023 Meehl, G. A., Covey, C., Delworth, T., Latif, M., Mcavaney, B., Mitchell, J. F. B., Stouffer, R. J. &
2222
2223
1024 Taylor, K. E. 2007. THE WCRP CMIP3 Multimodel Dataset: A New Era in Climate Change
2224
2225
1025 Research. Bulletin of the American Meteorological Society, 88, 1383-1394.
2226
2227 1026 Mein, R. G. & Larson, C. L. 1973. Modeling infiltration during a steady rain. Water Resources Research,
2228
2229 1027 9, 384-394.
2230
2231 1028 Mishra, A. K. & Singh, V. P. 2011. Drought modeling – A review. Journal of Hydrology, 403, 157-
2232
2233 1029 175.
2234
2235 1030 Mitchell, N., Kumarasamy, K., Cho, S. J., Belmont, P., Dalzell, B. & Gran, K. 2018. Reducing High
2236
2237 1031 Flows and Sediment Loading through Increased Water Storage in an Agricultural Watershed
2238
2239 1032 of the Upper Midwest, USA. Water, 10, 1053.
2240
2241 38
2242
2243
2244
2245
1033 Mohammed, K., Islam, A. K. M. S., Islam, G. M. T., Alfieri, L., Bala, S. K. & Khan, M. J. U. 2017a.
2246
2247
1034 Impact of High-End Climate Change on Floods and Low Flows of the Brahmaputra River.
2248
2249 1035 Journal of Hydrologic Engineering, 22, 04017041.
2250
2251 1036 Mohammed, K., Islam, A. K. M. S., Islam, G. M. T., Alfieri, L., Khan, M. J. U., Bala, S. K. & Das, M.
2252
2253 1037 K. 2018. Future Floods in Bangladesh under 1.5C, 2C, and 4C Global Warming Scenarios.
2254
2255 1038 Journal of Hydrologic Engineering, 23, 04018050.
2256
2257 1039 Mohammed, K., Islam, A. S., Islam, G. T., Alfieri, L., Bala, S. K. & Khan, M. J. U. 2017b. Extreme
2258
2259 1040 flows and water availability of the Brahmaputra River under 1.5 and 2 °C global warming
2260
2261 1041 scenarios. Climatic Change, 145, 159-175.
2262
2263 1042 Mohor, G. S. & Mendiondo, E. M. 2017. Economic indicators of hydrologic drought insurance under
2264
2265 1043 water demand and climate change scenarios in a Brazilian context. Ecological Economics, 140,
2266
2267 1044 66-78.
2268
2269 1045 Monteiro, J. a. F., Strauch, M., Srinivasan, R., Abbaspour, K. & Gücker, B. 2016. Accuracy of grid
2270
2271 1046 precipitation data for Brazil: application in river discharge modelling of the Tocantins
2272
2273 1047 catchment. Hydrological Processes, 30, 1419-1430.
2274
2275 1048 Moriasi, D. N., Arnold, J. G., Van Liew, M. W., Bingner, R. L., Harmel, R. D. & Veith, T. L. 2007.
2276
2277 1049 Model Evaluation Guidelines for Systematic Quantification of Accuracy in Watershed
2278
1050 Simulations. Transactions of the ASABE, 50, 885-900.
2279
2280
1051 Moriasi, D. N., Gitau, M. W., Pai, N. & Daggupati, P. 2015. Hydrologic and water quality models:
2281
2282
1052 Performance measures and evaluation criteria. Trans. ASABE, 58, 1763–1785.
2283
2284
1053 Nair, S. S. 2011. Importance of crop yield in calibrating watershed water quality simulation tools.
2285
2286 1054 Journal of the American Water Resources Association, v. 47, pp. 1285-1297-2011 v.47 no.6.
2287
2288 1055 Narasimhan, B. & Srinivasan, R. 2005. Development and evaluation of Soil Moisture Deficit Index
2289
2290 1056 (SMDI) and Evapotranspiration Deficit Index (ETDI) for agricultural drought monitoring.
2291
2292 1057 Agricultural and Forest Meteorology, 133, 69-88.
2293
2294 1058 Näschen, K., Diekkrüger, B., Leemhuis, C., Steinbach, S., Seregina, L. S., Thonfeld, F. & Van Der
2295
2296 1059 Linden, R. 2018. Hydrological Modeling in Data-Scarce Catchments: The Kilombero
2297
2298 1060 Floodplain in Tanzania. Water, 10, 599.
2299
2300 39
2301
2302
2303
2304
1061 Naveendrakumar, G., Vithanage, M., Kwon, H. H., Chandrasekara, S. S. K., Iqbal, M. C. M.,
2305
2306
1062 Pathmarajah, S., Fernando, W. & Obeysekera, J. 2019. South Asian perspective on temperature
2307
2308 1063 and rainfall extremes: A review. Atmospheric Research, 225, 110-120.
2309
2310 1064 Ncar 2019. APHRODITE: Asian Precipitation - Highly-Resolved Observational Data Integration
2311
2312 1065 Towards Evaluation) National Center for Atmospheric Research, University Corporation for
2313
2314 1066 Atmospheric Research, Boulder, CO. Available: https://climatedataguide.ucar.edu/climate-.
2315
2316 1067 Netwon-Nerc. 2019. Understanding impacts of flooding and drought in South East Asia [Online].
2317
2318 1068 Available: https://nerc.ukri.org/press/releases/2019/04-impacts/ [Accessed 30 August 2019].
2319
2320 1069 Nguyen-Tien, V., Elliott, R. J. R. & Strobl, E. A. 2018. Hydropower generation, flood control and dam
2321
2322 1070 cascades: A national assessment for Vietnam. Journal of Hydrology, 560, 109-126.
2323
2324 1071 Olivera, F., Valenzuela, M., Srinivasan, R., Choi, J., Cho, H., Koka, S. & Agrawal, A. 2006. ARCGIS-
2325
2326 1072 SWAT: A geodata model and GIS interface for SWAT1. JAWRA Journal of the American
2327
2328 1073 Water Resources Association, 42, 295-309.
2329
2330 1074 Perkins, S. P. & Sophocleous, M. 1999. Development of a Comprehensive Watershed Model Applied
2331
2332 1075 to Study Stream Yield under Drought Conditions. Groundwater, 37, 418-426.
2333
2334 1076 Pfannerstill, M., Guse, B. & Fohrer, N. 2014. Smart low flow signature metrics for an improved overall
2335
2336 1077 performance evaluation of hydrological models. Journal of Hydrology, 510, 447-458.
2337
1078 Piniewski, M., Szcześniak, M., Kundzewicz, Z. W., Mezghani, A. & Hov, Ø. 2017. Changes in low and
2338
2339
1079 high flows in the Vistula and the Odra basins: Model projections in the European-scale context.
2340
2341
1080 Hydrological Processes, 31, 2210-2225.
2342
2343
1081 Rahman, M., Bolisetti, T. & Balachandar, R. 2010. Effect of Climate Change on Low-Flow Conditions
2344
2345 1082 in the Ruscom River Watershed, Ontario. Transactions of the Asabe, 53, 1521-1532.
2346
2347 1083 Raikes, J., Smith, T. F., Jacobson, C. & Baldwin, C. 2019. Pre-disaster planning and preparedness for
2348
2349 1084 floods and droughts: A systematic review. International Journal of Disaster Risk Reduction,
2350
2351 1085 38, 101207.
2352
2353 1086 Rathjens, H., Katrin, B., Raghavan, S., Indrajeet, C. & Arnold, J. G. 2016. CMhyd User Manual.
2354
2355 1087 Richter, B. D., Baumgartner, J. V., Powell, J. & Braun, D. P. 1996. A Method for Assessing Hydrologic
2356
2357 1088 Alteration within Ecosystems. Conservation Biology, 10, 1163-1174.
2358
2359 40
2360
2361
2362
2363
1089 Richts, A. & Vrba, J. 2016. Groundwater resources and hydroclimatic extremes: mapping global
2364
2365
1090 groundwater vulnerability to floods and droughts. Environmental Earth Sciences, 75, 926.
2366
2367 1091 Righetto, J. M. & Mendiondo, E. M. 2007. Modelo de Seguro para Riscos Hidrológicos. Rev. Bras.
2368
2369 1092 Recur. Hídricos - RBRH 12, 107–113.
2370
2371 1093 Robi, M. A., Abebe, A. & Pingale, S. M. 2019. Flood hazard mapping under a climate change scenario
2372
2373 1094 in a Ribb catchment of Blue Nile River basin, Ethiopia. Applied Geomatics, 11, 147-160.
2374
2375 1095 Roth, V. & Lemann, T. 2016. Comparing CFSR and conventional weather data for discharge and soil
2376
2377 1096 loss modelling with SWAT in small catchments in the Ethiopian Highlands. Hydrol. Earth Syst.
2378
2379 1097 Sci., 20, 921-934.
2380
2381 1098 Ryu, J. H., Lee, J. H., Jeong, S., Park, S. K. & Han, K. 2011. The impacts of climate change on local
2382
2383 1099 hydrology and low flow frequency in the Geum River Basin, Korea. Hydrological Processes,
2384
2385 1100 25, 3437-3447.
2386
2387 1101 Schilling, K. E., Gassman, P. W., Kling, C. L., Campbell, T., Jha, M. K., Wolter, C. F. & Arnold, J. G.
2388
2389 1102 2014. The potential for agricultural land use change to reduce flood risk in a large watershed.
2390
2391 1103 Hydrological Processes, 28, 3314-3325.
2392
2393 1104 Sehgal, V. & Sridhar, V. 2018. Effect of hydroclimatological teleconnections on the watershed-scale
2394
2395 1105 drought predictability in the southeastern United States. International Journal of Climatology,
2396
1106 38, e1139-e1157.
2397
2398
1107 Seidou, O., Ramsay, A. & Nistor, I. 2012a. Climate change impacts on extreme floods I: combining
2399
2400
1108 imperfect deterministic simulations and non-stationary frequency analysis. Natural Hazards,
2401
2402
1109 61, 647-659.
2403
2404 1110 Seidou, O., Ramsay, A. & Nistor, I. 2012b. Climate change impacts on extreme floods II: improving
2405
2406 1111 flood future peaks simulation using non-stationary frequency analysis. Natural Hazards, 60,
2407
2408 1112 715-726.
2409
2410 1113 Senent-Aparicio, J., Jimeno-Sáez, P., Bueno-Crespo, A., Pérez-Sánchez, J. & Pulido-Velázquez, D.
2411
2412 1114 2019. Coupling machine-learning techniques with SWAT model for instantaneous peak flow
2413
2414 1115 prediction. Biosystems Engineering, 177, 67-77.
2415
2416
2417
2418 41
2419
2420
2421
2422
1116 Shepherd, B., Harper, D. & Millington, A. 1999. Modelling catchment-scale nutrient transport to
2423
2424
1117 watercourses in the U.K. Hydrobiologia, 395-396, 227-238.
2425
2426 1118 Shrestha, A., Sharma, S., Mclean, C. E., Kelly, B. A. & Martin, S. C. 2017. Scenario analysis for
2427
2428 1119 assessing the impact of hydraulic fracturing on stream low flows using the SWAT model.
2429
2430 1120 Hydrological Sciences Journal, 62, 849-861.
2431
2432 1121 Shrestha, S., Sharma, S., Gupta, R. & Bhattarai, R. 2019. Impact of global climate change on stream
2433
2434 1122 low flows: A case study of the great Miami river watershed, Ohio, USA. International Journal
2435
2436 1123 of Agricultural and Biological Engineering, 12, 84-95.
2437
2438 1124 Simons, M., Podger, G. & Cooke, R. 1996. IQQM—A hydrologic modelling tool for water resource
2439
2440 1125 and salinity management. Environmental Software, 11, 185-192.
2441
2442 1126 Singh, A. 2016. Modeling Stream Flow with Prediction Uncertainty by Using SWAT Hydrologic and
2443
2444 1127 RBNN Models for an Agricultural Watershed in India. National Academy Science Letters, 39,
2445
2446 1128 213-216.
2447
2448 1129 Singh, A., Imtiyaz, M., Isaac, R. K. & Denis, D. M. 2012. Comparison of soil and water assessment
2449
2450 1130 tool (SWAT) and multilayer perceptron (MLP) artificial neural network for predicting sediment
2451
2452 1131 yield in the Nagwa agricultural watershed in Jharkhand, India. Agricultural Water Management,
2453
2454 1132 104, 113-120.
2455
1133 Singh, J., Knapp, H. V., Arnold, J. G. & Demissie, M. 2005. Hydrological Modeling of the Iroquois
2456
2457
1134 River Watershed using HSPF And SWAT. JAWRA Journal of the American Water Resources
2458
2459
1135 Association, 41, 343-360.
2460
2461
1136 Singh, V. & Goyal, M. K. 2017. Unsteady High Velocity Flood Flows and the Development of Rating
2462
2463 1137 Curves in a Himalayan Basin under Climate Change Scenarios. Journal of Hydrologic
2464
2465 1138 Engineering, 22, 04017023.
2466
2467 1139 Singh, V., Goyal, M. K. & Chu, X. 2016. Multicriteria Evaluation Approach for Assessing Parametric
2468
2469 1140 Uncertainty during Extreme Peak and Low Flow Conditions over Snow Glaciated and Inland
2470
2471 1141 Catchments. Journal of Hydrologic Engineering, 21, 04015044.
2472
2473 1142 Singh, V. & Woolhiser, D. 2002. Mathematical Modeling of Watershed Hydrology. Journal of
2474
2475 1143 Hydrologic Engineering, 7, 270-292.
2476
2477 42
2478
2479
2480
2481
1144 Song, S., Schmalz, B. & Fohrer, N. 2014. Simulation and comparison of stream power in-channel and
2482
2483
1145 on the floodplain in a German lowland area. Journal of Hydrology and Hydromechanics 62,
2484
2485 1146 133.
2486
2487 1147 Spellman, P., Webster, V. & Watkins, D. 2018. Bias correcting instantaneous peak flows generated
2488
2489 1148 using a continuous, semi-distributed hydrologic model. Journal of Flood Risk Management,
2490
2491 1149 11, e12342.
2492
2493 1150 Stewart, I. T., Ficklin, D. L., Carrillo, C. A. & Mcintosh, R. 2015. 21st century increases in the
2494
2495 1151 likelihood of extreme hydrologic conditions for the mountainous basins of the Southwestern
2496
2497 1152 United States. Journal of Hydrology, 529, 340-353.
2498
2499 1153 Sun, Q., Miao, C., Duan, Q., Ashouri, H., Sorooshian, S. & Hsu, K.-L. 2018. A Review of Global
2500
2501 1154 Precipitation Data Sets: Data Sources, Estimation, and Intercomparisons. Reviews of
2502
2503 1155 Geophysics, 56, 79-107.
2504
2505 1156 SWAT 2019a. Soil and Water Assessment Tool: Software, ARCSWAT. Texas A&M University,
2506
2507 1157 College Station, TX and USDA-ARS Grassland, Soil and Water Research Laboratory, Temple,
2508
2509 1158 TX. Available at: https://swat.tamu.edu/software/arcswat/. [Accessed 15 October 2019]. .
2510
2511 1159 SWAT 2019b. Soil and Water Assessment Tool: Software, QSWAT. Texas A&M University, College
2512
2513 1160 Station, TX and USDA-ARS Grassland, Soil and Water Research Laboratory, Temple, TX.
2514
1161 Available at: https://swat.tamu.edu/software/qswat/. [Accessed 15 October 2019]. .
2515
2516
1162 SWAT 2019c. SWAT+. Texas A&M University, College Station, TX and USDA-ARS Grassland, Soil
2517
2518
1163 and Water Research Laboratory, Temple, TX. Available at:
2519
2520
1164 https://swat.tamu.edu/software/plus/. [Accessed 15 October 2019]. .
2521
2522 1165 Tan, M. L., Ficklin, D., Ibrahim, A. L. & Yusop, Z. 2014. Impacts and uncertainties of climate change
2523
2524 1166 on streamflow of the Johor River Basin, Malaysia using a CMIP5 General Circulation Model
2525
2526 1167 ensemble. Journal of Water and Climate Change, 5, 676–695.
2527
2528 1168 Tan, M. L., Gassman, P. W. & Cracknell, A. P. 2017. Assessment of Three Long-Term Gridded Climate
2529
2530 1169 Products for Hydro-Climatic Simulations in Tropical River Basins. Water, 9, 229.
2531
2532 1170 Tan, M. L., Gassman, P. W., Srinivasan, R., Arnold, J. G. & Yang, X. 2019a. A Review of SWAT
2533
2534 1171 Studies in Southeast Asia: Applications, Challenges and Future Directions. Water, 11, 914.
2535
2536 43
2537
2538
2539
2540
1172 Tan, M. L., Juneng, L., Tangang, F. T., Chan, N. W. & Ngai, S. T. 2019b. Future hydro-meteorological
2541
2542
1173 drought of the Johor River Basin, Malaysia, based on CORDEX-SEA projections. Hydrological
2543
2544 1174 Sciences Journal, 64, 921-933.
2545
2546 1175 Tan, M. L., Samat, N., Chan, N. W. & Roy, R. 2018. Hydro-Meteorological Assessment of Three GPM
2547
2548 1176 Satellite Precipitation Products in the Kelantan River Basin, Malaysia. Remote Sensing, 10,
2549
2550 1177 1011.
2551
2552 1178 Tan, M. L. & Santo, H. 2018. Comparison of GPM IMERG, TMPA 3B42 and PERSIANN-CDR
2553
2554 1179 satellite precipitation products over Malaysia. Atmospheric Research, 202, 63-76.
2555
2556 1180 Tan, M. L. & Yang, X. 2020. Effect of rainfall station density, distribution and missing values on SWAT
2557
2558 1181 outputs in tropical region. Journal of Hydrology, 584, 124660.
2559
2560 1182 Taylor, K. E., Stouffer, R. J. & Meehl, G. A. 2012. An Overview of CMIP5 and the Experiment Design.
2561
2562 1183 Bulletin of the American Meteorological Society, 93, 485-498.
2563
2564 1184 Tirupathi, C., Shashidhar, T. & Srinivasan, R. 2018. Analysis of rainfall extremes and water yield of
2565
2566 1185 Krishna river basin under future climate scenarios. Journal of Hydrology-Regional Studies, 19,
2567
2568 1186 287-306.
2569
2570 1187 Trudel, M., Doucet-Genereux, P. L. & Leconte, R. 2017. Assessing River Low-Flow Uncertainties
2571
2572 1188 Related to Hydrological Model Calibration and Structure under Climate Change Conditions.
2573
1189 Climate, 5, 24.
2574
2575
1190 Trung, L. D., Duc, N. A., Nguyen, L. T., Thai, T. H., Khan, A., Rautenstrauch, K. & Schmidt, C. 2018.
2576
2577
1191 Assessing cumulative impacts of the proposed Lower Mekong Basin hydropower cascade on
2578
2579
1192 the Mekong River floodplains and Delta – Overview of integrated modeling methods and
2580
2581 1193 results. Journal of Hydrology.
2582
2583 1194 Tuppad, P., Douglas-Mankin, K. R., Lee, T., Srinivasan, R. & Arnold, A. J. 2011. Soil and Water
2584
2585 1195 Assessment Tool (SWAT) hydrologic/water quality model: Extended capability and wider
2586
2587 1196 adoption. . Transactions of the ASABE, 54, 1677-1684.
2588
2589 1197 Tzoraki, O., Cooper, D., Kjeldsen, T., Nikolaidis, N. P., Gamvroudis, C., Froebrich, J., Querner, E.,
2590
2591 1198 Gallart, F. & Karalemas, N. 2013. Flood generation and classification of a semi-arid
2592
2593
2594
2595 44
2596
2597
2598
2599
1199 intermittent flow watershed: Evrotas river. International Journal of River Basin Management,
2600
2601
1200 11, 77-92.
2602
2603 1201 Ummenhofer, C. C. & Meehl, G. A. 2017. Extreme weather and climate events with ecological
2604
2605 1202 relevance: a review. Philosophical Transactions of the Royal Society B: Biological Sciences,
2606
2607 1203 372, 20160135.
2608
2609 1204 USACE U. a. C. O. E. 2013. HEC-ResSim Reservoir System Simulation User's Manual version 3.1.
2610
2611 1205 US Army Corps of Engineers, Institute for Water Resources. Hydrologic Engineering Center,
2612
2613 1206 609 Second Street Davis, CA 95616.
2614
2615 1207 USACE, U. a. C. O. E. 2016. HEC-RAS 5.0, User's Manual. Institute for Water Resources, Hydrologic
2616
2617 1208 Engineering Center, 609 Second Street, Davis, CA 95616.
2618
2619 1209 USDA-NRCS 2004a. Chapter 9: Hydrologic soil-cover complexes. National Engineering Handbook
2620
2621 1210 Hydrology Part 630. Washington, D.C.: U.S. Department of Agriculture, Natural Resources
2622
2623 1211 Conservation Service.
2624
2625 1212 USDA-NRCS 2004b. Chapter 10: Estimation of direct runoff from storm rainfall. National Engineering
2626
2627 1213 Handbook Hydrology Part 630. . Washington, D.C.: U.S. Department of Agriculture, Natural
2628
2629 1214 Resources Conservation Service.
2630
2631 1215 Van Griensven, A., Ndomba, P., Yalew, S. & Kilonzo, F. 2012. Critical review of SWAT applications
2632
1216 in the Upper Nile basin countries. Hydrological and Earth System Sciences, 16, 3371-3381.
2633
2634
1217 Van Liew, M. W., Garbrecht, J. D. & Arnold, J. G. 2003. Simulation of the impacts of flood retarding
2635
2636
1218 structures on streamflow for a watershed in southwestern Oklahoma under dry, average, and
2637
2638
1219 wet climatic conditions. Journal of Soil and Water Conservation, 58, 340-348.
2639
2640 1220 Verburg, P. H., Soepboer, W., Veldkamp, A., Limpiada, R., Espaldon, V. & Mastura, S. S. A. 2002.
2641
2642 1221 Modeling the Spatial Dynamics of Regional Land Use: The CLUE-S Model. Environmental
2643
2644 1222 Management, 30, 391-405.
2645
2646 1223 Vu, M. T., Raghavan, V. S. & Liong, S.-Y. 2015. Ensemble Climate Projection for Hydro-
2647
2648 1224 Meteorological Drought over a river basin in Central Highland, Vietnam. KSCE Journal of
2649
2650 1225 Civil Engineering, 19, 427-433.
2651
2652
2653
2654 45
2655
2656
2657
2658
1226 Walters, K. M. & Babbar-Sebens, M. 2016. Using climate change scenarios to evaluate future
2659
2660
1227 effectiveness of potential wetlands in mitigating high flows in a Midwestern U.S. watershed.
2661
2662 1228 Ecological Engineering, 89, 80-102.
2663
2664 1229 Wang, D., Hejazi, M., Cai, X. & Valocchi, A. J. 2011. Climate change impact on meteorological,
2665
2666 1230 agricultural, and hydrological drought in central Illinois. Water Resources Research, 47.
2667
2668 1231 Wang, R., Yuan, Y., Yen, H., Grieneisen, M., Arnold, J., Wang, D., Wang, C. & Zhang, M. 2019. A
2669
2670 1232 review of pesticide fate and transport simulation at watershed level using SWAT: Current status
2671
2672 1233 and research concerns. Science of The Total Environment, 669, 512-526.
2673
2674 1234 Waterloo, H. 2011. Visual MODFLOW version 2011.1. User manual. Waterloo Hydrogeologic Inc.,
2675
2676 1235 Waterloo.
2677
2678 1236 Wilhite, D. A. & Glantz, M. H. 1985. Understanding: the Drought Phenomenon: The Role of
2679
2680 1237 Definitions. Water International, 10, 111-120.
2681
2682 1238 Williams, J. R., Arnold, J. G., Kiniry, J. R., Gassman, P. W. & Green, C. H. 2008. History of model
2683
2684 1239 development at Temple, Texas. Hydrological Sciences Journal, 53, 948-960.
2685
2686 1240 Williams, J. R., Kannan, N., Wang, X., Santhi, C. & Arnold, J. G. 2012. Evolution of the SCS Runoff
2687
2688 1241 Curve Number Method and Its Application to Continuous Runoff Simulation. Journal of
2689
2690 1242 Hydrologic Engineering, 17, 1221-1229.
2691
1243 World Meteorological Organization (Wmo) & Global Water Partnership (Gwp) 2016. Handbook of
2692
2693
1244 Drought Indicators and Indices (M. Svoboda and B.A. Fuchs). Integrated Drought Management
2694
2695
1245 Programme (IDMP), Integrated Drought Management Tools and Guidelines Series 2. Geneva.
2696
2697
1246 Wu, J., Yen, H., Arnold, J. G., Yang, Y. C. E., Cai, X., White, M. J., Santhi, C., Miao, C. & Srinivasan,
2698
2699 1247 R. 2020. Development of reservoir operation functions in SWAT+ for national environmental
2700
2701 1248 assessments. Journal of Hydrology, 583, 124556.
2702
2703 1249 Xie, K., Liu, P., Zhang, J., Libera, D. A., Wang, G., Li, Z. & Wang, D. 2020. Verification of a New
2704
2705 1250 Spatial Distribution Function of Soil Water Storage Capacity Using Conceptual and SWAT
2706
2707 1251 Models. Journal of Hydrologic Engineering, 25, 04020001.
2708
2709
2710
2711
2712
2713 46
2714
2715
2716
2717
1252 Xu, X., Wang, Y.-C., Kalcic, M., Muenich, R. L., Yang, Y. C. E. & Scavia, D. 2017. Evaluating the
2718
2719
1253 impact of climate change on fluvial flood risk in a mixed-used watershed. Environmental
2720
2721 1254 Modelling & Software.
2722
2723 1255 Yaduvanshi, A., Sharma, R. K., Kar, S. C. & Sinha, A. K. 2018. Rainfall–runoff simulations of extreme
2724
2725 1256 monsoon rainfall events in a tropical river basin of India. Natural Hazards, 90, 843-861.
2726
2727 1257 Yan, D., Shi, X., Yang, Z., Li, Y., Zhao, K. & Yuan, Y. 2013. Modified Palmer Drought Severity Index
2728
2729 1258 Based on Distributed Hydrological Simulation. Mathematical Problems in Engineering, 2013,
2730
2731 1259 8.
2732
2733 1260 Yang, Y., Wang, G., Wang, L., Yu, J. & Xu, Z. 2014. Evaluation of Gridded Precipitation Data for
2734
2735 1261 Driving SWAT Model in Area Upstream of Three Gorges Reservoir. PLOS ONE, 9, e112725.
2736
2737 1262 Yatagai, A., Kamiguchi, K., Arakawa, O., Hamada, A., Yasutomi, N. & Kitoh, A. 2012. APHRODITE:
2738
2739 1263 Constructing a Long-Term Daily Gridded Precipitation Dataset for Asia Based on a Dense
2740
2741 1264 Network of Rain Gauges. Bulletin of the American Meteorological Society, 93, 1401-1415.
2742
2743 1265 Yu, D., Xie, P., Dong, X., Hu, X., Liu, J., Li, Y., Peng, T., Ma, H., Wang, K. & Xu, S. 2018a.
2744
2745 1266 Improvement of the SWAT model for event-based flood simulation on a sub-daily timescale.
2746
2747 1267 Hydrol. Earth Syst. Sci., 22, 5001-5019.
2748
2749 1268 Yu, Y., Liu, J., Yang, Z., Cao, Y., Chang, J. & Mei, C. 2018b. Effect of climate change on water
2750
1269 resources in the Yuanshui River Basin: a SWAT model assessment. Arabian Journal of
2751
2752
1270 Geosciences, 11, 270.
2753
2754
1271 Zhang, X., Booij, M. J. & Xu, Y.-P. 2015. Improved Simulation of Peak Flows under Climate Change:
2755
2756
1272 Postprocessing or Composite Objective Calibration? Journal of Hydrometeorology, 16, 2187-
2757
2758 1273 2208.
2759
2760 1274 Zhang, X., Xu, Y.-P. & Fu, G. 2014. Uncertainties in SWAT extreme flow simulation under climate
2761
2762 1275 change. Journal of Hydrology, 515, 205-222.
2763
2764 1276 Zhang, Y., Zhao, Y., Wang, Q., Wang, J., Li, H., Zhai, J., Zhu, Y. & Li, J. 2016. Impact of Land Use
2765
2766 1277 on Frequency of Floods in Yongding River Basin, China. Water, 8, 401.
2767
2768
2769
2770
2771
2772 47
2773
2774
2775
2776
1278 Zhao, P. P., Lu, H. S., Yang, H. C., Wang, W. C. & Fu, G. B. 2019. Impacts of climate change on
2777
2778
1279 hydrological droughts at basin scale: A case study of the Weihe River Basin, China. Quaternary
2779
2780 1280 International, 513, 37-46.
2781
2782 1281 Zou, L., Xia, J. & She, D. 2017. Drought Characteristic Analysis Based on an Improved PDSI in the
2783
2784 1282 Wei River Basin of China. Water, 9, 178.
2785
2786
2787 1283
2788
2789
1284
2790
2791
2792 1285
2793
2794
1286
2795
2796
2797 1287
2798
2799
2800 1288
2801
2802 1289
2803
2804
2805 1290
2806
2807 1291
2808
2809
2810 1292
2811
2812
1293
2813
2814
2815 1294
2816
2817
1295
2818
2819
2820 1296
2821
2822
2823 1297
2824
2825 1298
2826
2827
2828
2829
2830
2831 48
2832
2833
2834
2835
2836
1299 List of Figures
2837
2838
1300
2839
2840
2841
2842
2843
2844
2845
2846
2847
2848
2849
2850
2851
2852
2853
2854
2855
2856 1301
2857
1302 Figure 1: Common questions before, during and after SWAT hydro-climatic extreme studies.
2858
2859
1303
2860
2861
2862
2863
2864
2865
2866
2867
2868
2869
2870
2871
2872
2873
2874
2875
2876
2877
2878
2879
2880
2881
2882
2883 1304
2884
2885
1305 Figure 2: SWAT hydro-climatic extremes publications based on (a) publication years, (b) basin size
2886
1306 and (c) countries.
2887
2888
2889
2890 49
2891
2892
2893
2894
1307
2895
2896
2897
2898
2899
2900
2901
2902
2903
2904
2905
2906
2907 1308
2908
2909
1309 Figure 3: Finest streamflow calibration time-scale reported in the selected publications.
2910
1310
2911
2912
2913
2914
2915
2916
2917
2918
2919
2920
2921
2922
2923
2924
2925 1311
2926
2927 1312 Figure 4: Overall SWAT performance of the selected 111 publications.
2928
2929 1313
2930
2931 1314
2932
2933 1315
2934
2935 1316
2936
2937 1317
2938
2939 1318
2940
2941 1319
2942
2943 1320
2944
2945
1321
2946
1322
2947
2948
2949 50
2950
2951
2952
2953
2954
1323 List of Tables:
2955
2956
1324 Table 1: Selected publication for the SWAT hydro-climatic review.
2957 Climate
2958 Reference Region Focus Projection Special Notes / Keywords
2959 (Campbell et al., 2018) USA Assessment No Time-continuous, 5-min
(Kumar and Lakshmi, 2018) Transnational Assessment No Precipitation Intensities Analysis, TRMM
2960 (Li et al., 2018a) China Assessment No Event-based + FDC comparison
2961 (Spellman et al., 2018) USA Assessment No Peak flow comparison
(Tan et al., 2018) Malaysia Assessment No Event-based, GPM
2962 Event-based + Precipitation Intensities
2963 (Yaduvanshi et al., 2018) India Assessment No Comparison
2964 (Yu et al., 2018b) China Assessment No Event-based, 2-hour
(Bacopoulos et al., 2017) USA Assessment No Event-based, hourly
2965 (Boithias et al., 2017) France Assessment No Event-based, hourly
2966 EURO-
(Piniewski et al., 2017) Transnational Assessment CORDEX Peak & Low flows comparison
2967 (Dakhlalla and Parajuli, 2016) USA Assessment Synthetic Peak flow comparison
2968 (Singh et al., 2016) India Assessment No Peak & low flows comparison
2969 (Zhang et al., 2015) China Assessment CMIP3 Peak flow comparison
(Javaheri and Babbar-Sebens,
2970 2014) USA Assessment No Peak flow comparison
2971 (Pfannerstill et al., 2014) Germany Assessment No FDC comparison
(Dash et al., 2019) India Drought No Copula, Drought indices
2972 (Hoyos et al., 2019) Colombia Drought Synthetic Water yield changes
2973 (Shrestha et al., 2019) USA Drought CMIP5 FDC, 7Q10, 1Q10
2974 CORDEX-
(Tan et al., 2019b) Malaysia Drought SEA Drought Indices
2975 (Zhao et al., 2019) China Drought CMIP5 Drought Indices
2976 (Ahn et al., 2018) USA Drought No Irrigated agriculture, drought period
(Bayissa et al., 2018) Ethiopia Drought No Drought Indices
2977 (Lee et al., 2018) South Korea Drought CMIP5 Drought Indices
2978 (Li et al., 2018b) China Drought No Threshold value method
(Sehgal and Sridhar, 2018) USA Drought No Correlation Drought Indices
2979
(Chattopadhyay et al., 2017) USA Drought CMIP5 Drought indices
2980 (Kamali et al., 2017) Iran Drought ISI-MIP Drought Indices
2981 (Kang and Sridhar, 2017b) USA Drought CMIP5 Drought indices
(Kang and Sridhar, 2017a) USA Drought CMIP5 Drought Indices
2982 (Li et al., 2017) USA Drought No Ecosystem Services
2983 (Lweendo et al., 2017) Zambia Drought No Drought Indices
(Shrestha et al., 2017) USA Drought SDSM hydraulic fracturing, 7Q10
2984 (Zou et al., 2017) China Drought No SWAT_PDSI
2985 (Chen and Li, 2016) China Drought No Water scarcity
2986 (Brown et al., 2015) Australia Drought No FDC, drought period assesment
(Cai et al., 2015) USA Drought CMIP5 Drought prepareness
2987 (Jain et al., 2015) India Drought No Integrated Drought Vulnerability Index
2988 (Vu et al., 2015) Vietnam Drought CMIP3 Drought Indices
(Ashraf Vaghefi et al., 2014) Iran Drought CMIP3 critical continuous day calculator
2989 (Ryu et al., 2011) South Korea Drought CMIP3 7Q10
2990 (Wang et al., 2011) USA Drought CMIP3 Drought Indices
2991 (Rahman et al., 2010) Canada Drought CMIP3 FDC
(Narasimhan and Srinivasan,
2992 2005) USA Drought No Drought Indices
2993 (Igarashi et al., 2019) Thailand Flood Synthetic CLUEs, return period
(Jodar-Abellan et al., 2019) Spain Flood Synthetic Sub-daily SWAT, flash flood
2994 (Maghsood et al., 2019) Iran Flood CIMP5 Flood frequency Index, peak flow
2995 (Gao et al., 2018) China Flood Synthetic Copula, peak flow
2996 (Huang et al., 2018) China Flood No Partial Least Square Regression
(Iqbal et al., 2018) Transnational Flood CMIP5 HEC-SSP (flood frequency)
2997 CMIP3,
2998 (Kharel and Kirilenko, 2018) USA Flood CMIP5 Overspill risks
(Mitchell et al., 2018) USA Flood No Water retention site simulation
2999 (Mohammed et al., 2018) Transnational Flood CMIP5 Peak synchronization
3000 (Nguyen-Tien et al., 2018) Transnational Flood No Hydropower
(Cheng et al., 2017) USA Flood Synthetic Flood Hazard Index
3001 (Lee et al., 2017) South Korea Flood No Sangal's method
3002 Synthetic,
3003 CMIP3,
(Xu et al., 2017) USA Flood CMIP5 Fluvial flood risk
3004 (Walters and Babbar-Sebens,
3005 2016) USA Flood NARCCAP Wetland-based flood mitigation
(Zhang et al., 2016) China Flood No Non-point flood alleviation
3006
3007
3008 51
3009
3010
3011
3012 (Kharel and Kirilenko, 2015) USA Flood CMIP3 Overspill probability
3013 (Schilling et al., 2014) USA Flood No >24-foot flood stage = flood day
SWAT, HEC-HMS & EvroFloods
3014
(Angelidis et al., 2010) Transnational Flood No comparison
3015 (Kehew et al., 2010) Egypt Flood Synthetic Palaeoflood
3016 (Van Liew et al., 2003) USA Flood No Flood retarting structure
(Alodah and Seidou, 2019) Canada Flood & Drought CMIP5 Kernel Density Estimations
3017 (Chen et al., 2019a) China Flood & Drought CMIP5 Indicators of Hydrologic Alterations (IHA)
3018 (Leta and Bauwens, 2018) Belgium Flood & Drought CMIP3 WETSPRO
(Leta et al., 2018) USA Flood & Drought CMIP5 WETSPRO
3019 CORDEX
3020 (Näschen et al., 2018) Tanzania Flood & Drought Africa FDC
3021 (Tirupathi et al., 2018) India Flood & Drought CORDEX Water yield
(Ahn and Merwade, 2017) USA Flood & Drought No Copula
3022 (Lu et al., 2017) China Flood & Drought No Wet and dry years
3023 (Mohammed et al., 2017b) Transnational Flood & Drought CORDEX Probability Density Function
(Mohammed et al., 2017a) Transnational Flood & Drought CORDEX Annual max / min flow
3024 (Stewart et al., 2015) USA Flood & Drought CMIP3 Stream temperature
3025 (Robi et al., 2019) Ethiphio Coupling CMIP5 SWAT + MIKEFlood
3026 (Chinnasamy et al., 2018) India Coupling No SWAT + MODFLOW + HEC-RAS
(Chiogna et al., 2018) Italy Coupling No SWAT + SVM
3027 (Kang and Sridhar, 2018b) USA Coupling No SWAT & VIC
3028 (Kang and Sridhar, 2018a) USA Coupling CMIP5 SWAT & VIC
(Trung et al., 2018) Transnational Coupling No SWAT + MIKEBasin
3029 (Arunyanart et al., 2017) Thailand Coupling CMIP3 SWAT + HECRAS
3030 (Jamrussri and Toda, 2017) Thailand Coupling No SWAT + iRIC
3031 (Lopes et al., 2017) Portugal Coupling No SWAT + ELCIRC
(Mohor and Mendiondo, 2017) Brazil Coupling No SWAT + Vulnerability + Financial
3032 (Singh and Goyal, 2017) India Coupling No SWAT + Mike11
3033 (Ahn et al., 2016) South Korea Coupling CMIP5 SWAT + MODSIM
SWAT + Regional-scale Habitat
3034 (Esfahanian et al., 2016) USA Coupling No Suitability
3035 (Kharel et al., 2016) USA Coupling No SWAT + Economic model
3036 (Arias et al., 2014) Transnational Coupling No SWAT + IQQM + HecResSim + 2DEIA
(Gies et al., 2014) Transnational Coupling No SWAT + System dynamics
3037 (Song et al., 2014) Germany Coupling No SWAT + HECRAS
3038 (Kuntiyawichai et al., 2011) Thailand Coupling No SWAT + SOBEK
(Perkins and Sophocleous,
3039 1999) USA Coupling No SWAT + MODFLOW
3040 (Senent-Aparicio et al., 2019) Spain Improvement No Machine Learning
(Chilkoti et al., 2018) Canada Improvement No Multi-objective autocalibration
3041
(Duan et al., 2018) China Improvement No Modify Snow & Flood
3042 (Kim et al., 2018) South Korea Improvement No Bayesion Network
3043 (Ashraf Vaghefi et al., 2017) United States Improvement CMIP5 Climate Change Toolkit
(Esfahanian et al., 2017) USA Improvement No MASH Drought Index
3044 (Gharib et al., 2017) Canada Improvement No Extreme Value Analysis
3045 (McDaniel et al., 2017a) USA Improvement No SWAT + EPIC
(McDaniel et al., 2017b) USA Improvement No Crop-specific Drought Index
3046 (McDaniel et al., 2017c) USA Improvement No Crop-specific Drought Index
3047 (Trudel et al., 2017) Canada Improvement No Low flow uncertainty
3048 (Cohen Liechti et al., 2014) Transnational Improvement No Modify SWAT reservoir
(Leon et al., 2014) USA Improvement No Flood operation Model
3049 (Zhang et al., 2014) China Improvement PRECIS Uncertainty Analysis
3050 (Tzoraki et al., 2013) Greece Improvement No Flood Classification
(Yan et al., 2013) China Improvement No Modified PDSI
3051 (Seidou et al., 2012a) Canada Improvement No Improve peak flow
3052 (Seidou et al., 2012b) Canada Improvement CMIP3 Statistical downscaling
3053 1325
3054
3055 1326
3056
3057
3058
3059
3060
3061
3062
3063
3064
3065
3066
3067 52
3068
Declaration of interests

☒ The authors declare that they have no known competing financial interests or personal relationships
that could have appeared to influence the work reported in this paper.

☐The authors declare the following financial interests/personal relationships which may be considered
as potential competing interests:
Mou Leong Tan: Conceptualization, Methodology, Formal analysis, Investigation,

Funding acquisition, Writing – Original draft, review & editing, Project administration

Philip Gassman: Conceptualization, Funding acquisition, Supervision, Writing –

review & editing Funding acquisition

Xiaoying Yang: Funding acquisition, Writing – review & editing.

James Haywood: Funding acquisition, Writing – review & editing.