AGRICULTURAL DROUGHT MONITORING

AND ASSESSMENT

By: VIMAL SHUKLA MTECH (R.S& GIS) MANIT,BHOPAL

Drought is a normal phenomenon of earths

climate, and a common feature in drylands

Drought severity & recurrence is maximum in arid regions

Compels Govt to spend huge amount in relief and rehabilitation Yet, lack of proper assessment and warning systems lead to confusion or delay in reaching affected people/region Needs development of robust assessment and monitoring tools Indicators are available for measuring: Meteorological, Hydrological and Agricultural drought Socio-economic Indicators are few

None of the current indicators are universally acceptable Also, not all is suitable for every region

Arid regions need a set of robust indicators and assessment tools for their high vulnerability

Physical, Biological and Social Indicators

Physical indicators include

Rainfall, Effective soil moisture, Surface water availability, Depth to groundwater, etc.

Biological/ Agricultural indicators comprise

Vegetation cover & composition, Crop & Fodder yield, Condition of domestic animals, Pest incidence, etc.

Social indicators are mostly impact indicators and include

Food and Feed availability, Land use conditions, Livelihood shifts, Migration of population, etc.

In most cases only those indicators that measure the rainfall needs of following sectors are considered: (a) agricultural need, (b) drinking water supply, and (c) storage of reservoirs and ground water

900 800 700

y = 0.1799x + 324.54 R2 = 0.0022

Annual rainfall (mm)

600 500 400 300 200 100 0

1901

1906

1911

1916

1921

1926

1931

1936

1941

1946

1951

1956

1961

1966

1971

1976

1981

1986

1991

1996

2001
3
0

Year

Long-term Rainfall trend of Arid Rajasthan

Annual rainfall (mm)

500 450 400 350 300 250 200 150 100 50 0

Rainy days 406.7

6 5

454

Freq. of flood Frequ. of drought 4

328

305.1

286.6
3 2

1 0

18
1960-69

25
1970-79

17
1980-89

21

18

0
0

1990-1999 2000-2007

1960-69

1970-79

1980-89

1990-1999

2000-2007

2006

Decade-wise Meteorological Drought in Arid Rajasthan

4
Frequency (Years)

R2 = 0.3278

R2 = 0.4914

R2 = 0.4683

0
19011910 19111920 19211930 19311940 19411950 19511960 19611970 19711980 19811990 19912000 20012005

Decade

Moderate Poly. (Moderate)

Severe Poly. (Severe)

Very Severe Poly. (Very Severe)

METEOROLOGICAL DROUGHT INDICATORS

Deciles of Precipitation (DI)
Monthly precipitation totals from a long-term record (~30 years) are used for deciles, which are grouped further into five precipitation classes : 0-20% 20 to 40% 40 to 60% 60 to 80% much below normal below normal near-normal above-normal

80 to 100%

much above normal

DI is used widely in Australia for drought relief programme. Merit : Demerit: DI is simple to calculate, requires only precipitation data and fewer assumptions. Too simplistic to inform about gravity of the problem in different sectors.

Precipitation Departure from Normal

IMD describes meteorological drought from rainfall departure from its long term averages and declares meteorological drought on weekly/monthly basis. Departure of annual rainfall from normal (%) 0 or above No drought 0 to 25 Mild drought -26 to 50 Moderate drought -50 or more Severe drought When >50% area of the country gets moderate or severe drought, the country becomes severely drought-affected; if 26-50% area is affected, country becomes moderately drought-affected. Merit : Simplicity makes this index popular in India.

Demerit : Average precipitation is not always the same as median precipitation. Also, distribution or time-scale of rainfall is not specified.

Palmer Drought Severity Index (PDSI)

PDSI, popular in the US, uses data on precipitation, temperature and local available water content (AWC) of soil, and calculates the difference between Climatically Appropriate For Existing Conditions (CAFEC) rainfall and actual rainfall as a drought indicator. PDSI generally varies between -4.0 (extreme drought) and +4.0 (adequate moisture condition). Drought categories are: Index value Class for drought - 1.00 to 1.99 Mild drought - 2.00 to 2.99 Moderate drought - 3.00 to - 3.99 Severe drought < - 4.00 Extreme drought Merit : PDSI quantifies abnormality of weather in a region, including in historical past. It can well be used for spatio-temporal variability of drought. Demerit : The index values did not often match the situation in India.

35

30

Palmer's Drought Severity Index for Western Rajasthan (Long-term, Monthly)

Mild Moderate Severe Extreme

Percentage frequency

25

20

15

10

June

July

August

September

Standardized Precipitation Index (SPI)

SPI, based on probability of precipitation for any time scale, is calculated as : X -Xm SPI = ---------- Where X = Precipitation for the station Xm = Mean precipitation = Standardized deviation SPI Drought Classes Less than -2.00 Extreme drought -1.50 to -1.99 Severe drought -1.00 to -1.49 Moderate drought -0.99 to -0.00 Mild drought Merits : Can be computed for different time scales Can provide early warning of drought Can help assess drought severity Is less complex than PDSI Groundwater, stream flow, and reservoir storage reflect longer-term precipitation anomalies. So, SPI is calculated for 3, 6, 12, 24, and 48 month time scales.

Demerits:

Relationship between Standardized Precipitation Index and Pearl millet yield of Jodhpur district (1971-2006)
800 700 600 500 400 300 200 100 0 -0.50 0.00
Pearl millet grain yield (kg/ha)

-2.00

-1.50

-1.00

0.50

1.00

1.50

2.00

2.50

Standardized Precipitation Index

SPI AND PEARL MILLET YIELD (Kg/ha) IN WESTERN RAJASTHAN

224

76 18

1137 946

152
126 70 7 158 0

118
286 583

986

866

1404

1051 1218 1166 552 1259 753

197

VALUES WITHIN DISTRICTS ARE AVERAGE PEARL MILLET YIELD (Kg/ ha)

HYDROLOGICAL DROUGHT INDICATORS

Groundwater and Reservoir Level Monitoring of all reservoir water levels and groundwater table through a closed well observation network is important. Standardized Water level Index (SWI) An index based on water level probability for any time scale. SWI= (Wij Wim)/ where, Wij is the seasonal water level for ith and jth observation, Wim its seasonal mean, and is its standard deviation. Merits: SWI can be computed for different time scales Can provide early warning of water storage Can help in assessing hydrological drought severity

Surface Water Supply Index (SWSI)

Designed for river basins with a component of mountain snow input. Integrates reservoir storage, stream flow and snow and rain into a single index.

where a, b, c, and d are weights for snow, rain, stream flow and reservoir storage, respectively; while (a+b+c+d) = 1, and Pi = probability (%) of non-exceedence for each of the four water balance components. Calculated at monthly time step. Demerits : Unique to each basin or region, so difficult to compare across basins or regions. Changes in water management in a basin, necessitates redevelopment of the algorithm. Extreme events cause a problem if events surpass historical time series.

Reclamation Drought Index (RDI)

RDI is calculated at river basin level.
Inputs: temperature, precipitation, snow pack, stream flow, reservoir level. Impetus came from the Reclamation States Drought Assistance Act of 1988 in the USA, for seeking drought assistance.

RDI Classification
4.0 or more Extremely wet 1.0 to 1.5 Normal to mild wet 1.5 to 4.0 Moderately wet 0.0 to -1.5 Normal to mild drought

-1.5 to -4.0

Moderate drought

-4.0 or less

Extreme drought

RDI is similar to SPI, PDSI, and SWSI.

Merit : Builds a temperature-based demand component and a

duration into the index. Can account for both climate and water supply factors.

Demerit : Index is unique to each river basin, so inter-basin

comparison is limited.

AGRICULTURAL DROUGHT INDICATORS Aridity Index

Aridity Index indicates water-deficit conditions in a region. Crop-water requirements are not considered. Calculated as percentage ratio of annual water deficit to annual water need or annual potential evapotranspiration (PE). Aridity anomaly (Ia) index is the percent departure of the anomaly value from the normal. IMD monitors Ia during kharif season for the country as a whole and during rabi season for areas receiving NE monsoon rains. Drought Category Mild drought Moderate drought Severe drought Anomaly Value Up to 25% 26-50% > 50%

Demerit: Although simple, water balance calculations do not properly account for rainfall-runoff before the stored moisture is estimated.

ARIDITY INDEX FOR INDIA 16-29 JULY 2007

IMD, Pune

Moisture Adequacy Index (MAI)

CAZRI developed MAI for quantification of agricultural drought, which is defined as :

MAI = AE/PE
where AE is actual evaporation, and PE potential evapo-transpiration (in %) during different phonological stages of a crop. MAI is obtained from weekly water balance. Drought impact is related to moisture availability at certain crop growth stages. Hence, categories of MAI (severity) at different growth stages are integrated into a single index value to identify drought impact on a particular crop. Merit:

Water balance calculation takes into account soil characteristic, crop growth period and water requirement of major crops. Drought is specified crop-wise on a realtime basis. implement under data-scarce conditions.

Demerit: Calculations are data-intensive, and hence difficult to

Agricultural Drought Code Developed by CAZRI (Based on Moisture Adequacy Index)

AE/PE (%) during different phenophases
76 to 100 51 to 75 26 to 50 25 or less

Drought intensity

Phenophasewise Code
Seedling (S) S0 S1 S2 S3 Vegetative (V) V0 V1 V2 V3 Reproductive (R) R0 R1 R2 R3

No drought Mild Drought Moderate drought Severe drought

Intensity of Agricultural Drought

Sr.No. 1 2 3 No drought S0 V0 R0 S0 V0 R1 S0 V1 R0 Mild S0 V0 R2 S0 V1 R2 S0 V2 R0 Moderate S0 V0 R3 S0 V1 R3 S0 V2 R2 Severe S0 V2 R3 S0 V3 R2 S0 V3 R3

4
5 6 7

S1 V0 R0
S1 V0 R1 S1 V1 R0 S0 V1 R1

S0 V2 R1
S1 V0 R2 S1 V1 R1 S1 V1 R2

S0 V3 R0
S0 V3 R1 S1 V0 R3 S1 V1 R3

S1 V2 R3
S1 V3 R2 S1 V3 R3 S2 V0 R3

Based on the criteria developed by CAZRI, western Rajasthan can be broadly divided into three major zones of agricultural drought:

(1) Zone of Maximum Occurrence (1 in 2-3 Y);

(2) Zone of Medium Occurrence (1 in 3-4 Y); (3) Zone of Fewer Occurrence (1 in 4-5 Y)

Crop Water Stress Index (CWSI)

CWSI values are a daily integration of plant-available soil water, evaporative demand and plant phenological stage susceptibility, and is defined for the growing season as:
Harvest

CWSI= (1-(T/Tp) SUS

Planting

where, T is the computed actual transpiration (mm/day), Tp is potential transpiration (mm/day) and SUS is seasonally dependent weighting factor for grain yield susceptibility. SPAW model is used for simulation of soil water and calculation of effective rainfall for plant transpiration.

Merits : The estimates using dynamic simulation models are reasonably good.
Demerits: SPAW model needs calibration for each crop and region and hence has a limitation for use.

5000 4500
Pearl millet grain yield (kg/ ha)

4000 3500 3000 2500 2000 1500 1000 500 0 0 2 4 6 8 10 12 14

Water Stress Index (SPAW model)

Relationships between Water Stress Index and Pearl millet Yield

Comparative Performance of Agricultural Drought Indicators in Jodhpur District

Indicator Correlation with pearl millet yield (1971-2005) R2 value Merits and demerits

Standardized Precipitation Index Aridity Index Palmers Drought Severity Index Moisture Adequacy Index

0.216 0.380 0.473 0.673

Can be computed for different time scales, is less complex than the PDSI Water balance calculations have limitations for not accounting rainfall-runoff Meteorological index, responds better to abnormally dry or wet weather Can specify drought crop-wise and on real-time basis

Crop Water Stress Index (SPAW)

0.890

Simulates soil-water and calculates effective rainfall for plant transpiration, with reasonable results. Model needs calibration for each crop and region

DROUGHT-RELATED INDICES FROM REMOTE SENSING

Major Indices: NDVI, EVI, VCI, TVI, etc. Normalized Difference Vegetation Index (NDVI)

where is reflectance in the near-infra-red (NIR) and red (red) band of satellite sensor, respectively. NDVI ranges from -1 to 1.
Drought severity is evaluated as difference between NDVI for current month (e.g. September 2007) and a long-term (30-year-long) mean NDVI for the month. Since 1989 NADAMS is providing bi-weekly drought bulletins for kharif season at district level in India, based on satellite-derived greenness of plant cover. Merits : Calculation simple; daily satellite data available; several sensor wavelengths and calculation options now available

Demerits : Persistent cloud cover during monsoon Misrepresentation in sparsely vegetated areas Often lagging actual occurrence by weeks to month Does not yet reliably quantify biomass, crop condition, grain yield or even plant density CAZRIs experience with PD-54 index (Australia) for rangeland vegetation was fruitful than NDVI. The PD-54 was later improved as SAVI (soil adjusted vegetation index) and later modified (MSAVI).

Experience from CAZRIs Weather-based Agro-advisory Services to Farmers

Impact of agro-met advisory services in pearl millet and clusterbean cultivation Treatment No. of Grain yield Gross return Cultivation cost Net return B:C -1 -1 -1 -1 farmer (kg ha ) (Rs ha ) (Rs ha ) (Rs ha ) ratio Pearl millet AAS 39 950 4976 1569 3408 2.17 Non-AAS 37 725 3708 1328 2382 1.80 t-test value 2.00** Clusterbean AAS 13 700 5287 912 4374 4.80 Non-AAS 13 625 4440 1044 3398 3.25 t-test value 2.22** **Significant at 5% probability level (at P=0.01). AAS: Group of farmers following agro-met advisory services. Non-AAS: Group of farmers who did not follow agro-met advisory services.

Impact of agro-met advisory services in mustard and cumin cultivation

Treatment No. of Grain yield Gross return Cultivation cost Net return -1 -1 -1 -1 farmer (kg ha ) (Rs ha ) (Rs ha ) (Rs ha ) B:C Ratio Mustard AAS 26 1725 10743 5600 5143 0.92 Non-AAS 21 1650 10037 5789 4248 0.73 t-test value 2.07** Cumin AAS 24 450 10433 5229 5204 0.99 Non-AAS 17 275 6.98 5367 731 0.14 t-test value 2.00** **Significant at 1% probability level (at P=0.01). AAS: Group of farmers following agro-met advisory services. Non-AAS: Group of farmers who did not follow agro-met advisory services.

INDICATORS FOR DROUGHT EWS AND FOOD SECURITY, ESPECIALLY FOR AFRICA
Indicator Food crop performance Crop conditions Crop production forecast Marketing and price information Food supply/demand Health conditions Food crops and shortages Food supply Food consumption Crop areas Pests Food balance Vegetation front CCD NDVI Biomass Seeding risk areas Expected season length Estimated seeded areas Estimated seeding date Vegetation cover Agro-ecological zones Crop use intensity CV of agricultural production Cash crop production area Coping strategies Av. Travel cost to nearest market Livestock production Population density Access to water Childrens education Rainfall AP3A FIVIMS GIEWS SADC FEWS VAM

AP3A by AGRHYMET; FIVIMS by FAO; GIEWS by FAO; SADC by Zimbabwe; FEWS by USAID; VAM by WFP

NEED FOR A GIS-BASED DECISION SUPPORT SYSTEM TO MEASURE, MONITOR, WARN ABOUT AND MANAGE DROUGHT
Drought has DIFFERENT CONNOTATIONS and CONTEXTUAL RELATIONS in different areas and societal segments of India Consequently, NO ONE SET OF INDICES may provide a full glimpse of the problem in the country as a whole

Largest segment of society affected by drought in India depends on AGRICULTURE

Within agriculture sector, CROP CULTIVATORS ARE MOST VULNERABLE CAZRIs STUDIES on IMPACT OF 2002 DROUGHT showed that MEDIUM & SMALL FARMERS have MAXIMUM VULNERABILITY Marginal Farmers and affected Weaker Section of the society GET PRIORITY in SOCIAL SECURITY COVER during drought relief Next most-vulnerable appeared to be the LIVESTOCK RAISERS who migrate with large herds of animals DOMINANTLY RAIN-FED AGRICULTURE VILLAGES WITH POOR ACCESS to roads and other infrastructures constitute the MOST VULNERABLE AREAS TRADITIONAL WISDOM IN DROUGHT MANAGEMENT is getting eroded due to OVER-DEPENDENCE ON DROUGHT RELIEF

MINIMUM DATA LAYERS FOR A DSS ON DROUGHT VULNERABILITY & MONITORING

CLIMATE
RAINFALL TEMPERATURE

IN ARID AREAS LAND RESOURCES SOCIO-ECONOMIC LANDFORM SOIL TEXTURE SOIL DEPTH SOIL MOISTURE TREE/SHRUB COVER POPULATION STRUCTURE OCCUPATION STRUCTURE INFRASTRUCTURE LIVESTOCK COMPOSITION MARKET ACCESS

CROPS GROWN
IRRIGATION

TRANSPORT NETWORK
WATER AVAILABILITY
DATA SOURCES

REAL-TIME GROUND INFORMATION & SATELLITE PRODUCTS (incl. MICROWAVE), SECONDARY INFORMATION, SAMPLE SURVEY SOME KEY WORDS FOR MODELLING MOISTURE AVAILABILITY, FOOD AND FEED AVAILABILITY, VULNERABLE AREAS & GROUPS, DRINKING WATER, HUMAN & LIVESTOCK HEALTH, MIGRATION ROUTES, LIVELIHOOD OPTIONS, VILLAGE ACCESSIBILITY, TRADITIONAL ASSET CONDITION

A BARE MINIMUM DROUGHT MONITORING PLAN

Step 1: Collection of rainfall and temperature data from different locations (In collaboration with IMD and State agencies) Step 2: Calculation of temporal and spatial availability of soil moisture in a GIS environment using pre-calibrated dynamic simulation models for all major crops taking into water requirement and soil characteristics. Step: To find out threshold limits for each crop as a warning as no drought or mild, moderate and severe drought conditions for each location. Step 4: Taking medium range forecasting, preparation of early drought warning bulletins on drought status and disseminate to drought managers.

Step 5: Preparation of Agro-advisory bulletins based on drought conditions and contingency plans in case of late onset of monsoon.
Step 6: Dissemination of Agro-advisory bulletins to farmers through local media and get feed back.