Decadal Land Use Change and Regional climate modeling Experiments Applied to Water Management: West Central Florida

Case study
Jose L Hernandez, Syewoon Hwang, Guillermo Baigorria, Keith Ingram, James Jones, Wendy Graham
Agricultural and Biological Engineering University of Florida

Motivation
Climate and its analytical tools has a significant role in decision support systems for assisting operational management of water-resources. In the 20th century, Florida faced extensive transformation of land surface coverage due to agriculture, urbanization and alteration of water features ((1900-93) Marshal et al, 2004). Coupled regional climate models (RCM) present an enhanced tool helping:
to handle operational challenges in planning and managing sources of production (ground water, desalinization, stream flows); to assess environmental impact of urbanization and construction; to quantify uncertainties in budget and price of operation of WRM.

Agricultural and Biological Engineering University of Florida

Decision-Making
Fundamental question: Do we have water to sustain the demand of water by population, industrial and economical sectors at city, county or regional scales in a timeframe of 1 year to 18 months? Socio-economical models estimate long-term and relative large scale demands So, next season what could be the allocation demand at different spatial scales by sectors (population, agriculture, construction, industrial)? Allocation depends on availability of sources (in Florida: ground and surface water, desalinization) and 1 to 3 month predictions (CLIMATE) are needed to update information and refine decisions. Criteria for evaluation and decision: operational cost, reliability, climate concerns and environmental regulations. OUR GOAL IS TO PROVIDE A 1-3 Month FORECAST USING DYNAMICAL RCMs TO SUPPORT DECISION-MAKING IN WRM FOR C. FLORIDA .
Agricultural and Biological Engineering University of Florida

Area of each landuse (ha) 10

5 6 .5 6 1 .1 6 8 .3 7 0 .6 7 7 .8 7 5.9 7 0 .1 6 9 .4 1 6 .3 1 4 .6 1 3 .0 1 2 .7

x10000 90

20

30

40

50

60

70

80

0 Urban Agriculture Rangeland 7 Upland forests Water Wetland Barren land Transp. & util.
0 .4 0.6 1 .1 0 .6 3 .0 3.3 3 .7 3 .8 4 3 .2 4 1.6 4 0 .2 3 9 .1 4 4 .1 4 4.4 4 5 .4 4 5 .6 4 8 .7 4 8 .6 4 8 .3 4 8 .4

1995-2006 LU West Central Florida

Landuse categories

1995 1999 2004 2006

1995 landuse (%)

LU and Regional Climate modeling

Barren land, 0.1 Wetland, 17.9 Transp. & util., , 1.2 Urban, 21.8

Water, 7.6

-RCM use land surface models to describe energy and mass exchange of properties between land and atmosphere.
Upland forests, 16.2 Rangeland, 6.0 Agriculture, 29.3

1999 landuse (%)

Barren land, 0.2 Wetland, 16.8 Transp. & util., 1.1 Urban, 21.1

Water, 15.3

Upland forests, 14.3

Rangeland, 5.0

2004 landuse (%) Agriculture, 26.2

-A proper identification of land cover type and assignment of their physical properties are important in model performance describing temperature, sensible and latent heat flux, low atmosphere circulation, precipitation.
Barren land, 0.4 Transp. & util., 1.3 Urban, 23.5 2006 landuse (%)

Wetland, 16.7 Water, 15.6

Barren land, 0.2 Wetland, 16.7 Water, 15.7

Upland forests, 13.9 Agriculture, 24.1

Transp. & util., 1.3 Urban, 24.3

Rangeland, 4.5

Agricultural and Biological Engineering University of Florida

Upland forests, 13.5

Rangeland, 4.4

Agriculture, 23.9

Initial RCM Experiments and Model configurations

The PSU-NCAR MM5 version 3.7 Modeling system 1992-1993 USGS (MM5 LU model default data) 1986-2006 NCEP Reanalysis data (input for MM5 model simulations) Sensitivity experiments at 27Km, 9km and 3Km spatial resolutions , 21 sigma levels between 100K 40K Pa, under three physics configurations. University of Florida High Performance Computational System. Every region has its own controls on climate, and to select a better set of physics option for Florida we carry out sensibility experiments.
EXP# 1 2 3 FRAD 2 2 2 ICUPA 3 3 6 IMPHYS 4 6 4 IBLTYP 5 5 5 ISOIL 1 1 1

Agricultural and Biological Engineering University of Florida

Model and Observational data

LU Southwest Water Management District 1992-2005 FSU/COAPS meteor. data
2006 LU/SWMD

Three Model domains and Observations

36

34

x COAPS + FAWN

32

Latitude( C)

30

28

26

Dom3: 3 Km Dom2: 9 Km

Florida Automated Weather Network

24

22

Dom1: 27 Km
-90 -88 -86 -84 -82 -80 -78 -76 -74

-92

Agricultural and Biological Engineering University of Florida

Longitude(o)

Sensitivity Analysis: Temperature (D3: 3Km)

Exp. 1: Default LU
Ona(303) Temp 2m (oC) Temp 2m (oC) 35 30 25 20 2 4 6 8 10 Day in September 2005 Arcadia(320) 12 14 30 25 20 2 4 6 8 10 Day in September 2005 Carrabelle(310) 12 14 Balm(304) Temp 2m (oC) 35 30 25 20 2 4 6 8 10 Day in September 2005 Carrabelle(310) 12 14

Exp. 1: Modified LU
Ona(303) Temp 2m (oC) 35 30 25 20 2 4 6 8 10 Day in September 2005 Arcadia(320) 12 14 Balm(304)

Temp 2m (oC)

Temp 2m (oC)

Temp 2m (oC)

30 25 20 2 4 6 8 10 Day in September 2005 Live Oak(330) 12 14

30 25 20 2 4 6 8 10 Day in September 2005 Live Oak(330) 12 14

6 8 10 Day in September 2005 Lake Alfred(340)

12

14

Temp 2m (oC)

32 30 28 26 24 22

34 32 30 28 26 24 22 2 4 6 8 10 Day in September 2005 Lake Alfred(340) 12 14

34 32 30 28 26 24 22 2 4 6 8 10 Day in September 2005 Belle Glade(350) 12 14

Temp 2m (oC)

Temp 2m (oC)

Temp 2m (oC)

30 25 20 2 4 6 8 10 Day in September 2005 Fort Lauderdale(360) 12 14

Temp 2m (oC) 2 4 6 8 10 Day in September 2005 Belle Glade(350) 12 14

35 30 25

35 30 25 20 2 4 6 8 10 Day in September 2005 Fort Lauderdale(360) 12 14

Temp 2m (oC)

Temp 2m (oC)

Temp 2m (oC)

30 25 20 2 4 6 8 10 Day in September 2005 12 14

30 25 20 2 4 6 8 10 Day in September 2005 12 14

Temp 2m (oC)

34 32 30 28 26 24 22 2 4 6 8 10 Day in September 2005 12 14

35 30 25 2 4 6 8 10 Day in September 2005 12 14

StatID AveMod STDMod Ave_obs STD_obs Cor.Cf. RMSE 303 27.23 2.93 26.13 3.27 0.68 2.70 304 27.27 2.92 26.19 3.77 0.57 3.32 310 27.22 3.11 25.57 4.16 0.63 3.59 320 27.35 2.75 26.35 3.22 0.64 2.72 330 27.46 3.30 26.32 3.30 0.64 2.98 340 27.76 2.98 26.01 3.98 0.61 3.59 350 27.33 3.34 26.44 3.90 0.60 3.36 360 27.40 3.36 26.47 3.60 0.61 3.22

StatID AveMod STDMod AveObs 303 28.07 3.41 26.13 304 28.13 3.45 26.19 310 27.52 3.41 25.57 320 27.99 3.20 26.35 330 28.29 3.81 26.32 340 27.88 3.01 26.01 350 27.40 3.64 26.44 360 27.67 3.70 26.47

STDObs Cor.Coef. RMSE 3.27 0.71 3.19 3.77 0.60 3.77 4.16 0.63 3.80 3.22 0.65 3.09 3.30 0.65 3.52 3.98 0.60 3.69 3.90 0.61 3.47 3.60 0.60 3.45

Agricultural and Biological Engineering University of Florida

Sensitivity Analysis: Wind Speed Experiment # 1; 3 Km res.

Default LU
Ona(303) Wind Sp (m/s) Wind Sp (m/s) 6 4 2 2 4 6 8 10 Day in September 2005 Carrabelle(310) 12 14 6 4 2 2 4 6 8 10 Day in September 2005 Arcadia(320) 12 14 Balm(304)
Wind Sp (m/s) 6 4 2 2 4 6 8 10 Day in September 2005 Carrabelle(310) 12 14 Ona(303) Wind Sp (m/s) 6 4 2 2 4 6 8 10 Day in September 2005 Arcadia(320) 12 14

Modified LU
Balm(304)

Wind Sp (m/s)

Wind Sp (m/s)

Wind Sp (m/s)

6 4 2 2 4 6 8 10 Day in September 2005 Live Oak(330) 12 14

6 4 2 2 4 6 8 10 Day in September 2005 Live Oak(330) 12 14

Wind Sp (m/s)

6 4 2 2 4 6 8 10 Day in September 2005 Lake Alfred(340) 12 14

6 4 2 2 4 6 8 10 Day in September 2005 Lake Alfred(340) 12 14

Wind Sp (m/s)

Wind Sp (m/s)

Wind Sp (m/s)

5 4 3 2 1 2 4 6 8 10 Day in September 2005 Belle Glade(350) 12 14

5 4 3 2 1 2 4 6 8 10 Day in September 2005 Fort Lauderdale(360) 12 14

5 4 3 2 1 2 4 6 8 10 Day in September 2005 Belle Glade(350) 12 14

Wind Sp (m/s)

5 4 3 2 1 2 4 6 8 10 Day in September 2005 Fort Lauderdale(360) 12 14

Wind Sp (m/s)

Wind Sp (m/s) 2 4 6 8 10 Day in September 2005 12 14

Wind Sp (m/s)

Wind Sp (m/s)

6 4 2 2 4 6 8 10 Day in September 2005 12 14

6 4 2 2 4 6 8 10 Day in September 2005 12 14

6 4 2

4 3 2 1 2 4 6 8 10 Day in September 2005 12 14

StatID AveMod STDMod AveObs STDObs 303 2.94 1.28 2.69 1.33 304 2.93 1.26 3.01 1.58 310 2.62 1.15 3.04 1.61 320 2.91 1.27 3.16 1.44 330 2.77 1.09 2.15 1.06 340 2.59 1.31 2.10 1.33 350 2.73 1.20 3.10 1.58 360 2.60 1.21 1.55 0.96

Cor.Coef. 0.15 0.23 0.35 0.04 0.13 0.24 0.17 0.21

RMSE 1.71 1.78 1.67 1.89 1.47 1.68 1.80 1.68

StatID 303 304 310 320 330 340 350 360

AveMod STDMod AveObs 2.37 0.97 2.69 2.30 0.93 3.01 2.29 1.01 3.04 2.36 1.00 3.16 2.23 0.86 2.15 2.25 1.11 2.10 2.27 1.04 3.10 2.14 1.01 1.55

STDObs 1.33 1.58 1.61 1.44 1.06 1.33 1.58 0.96

Cor.Coef. RMSE 0.13 1.62 0.11 1.88 0.28 1.81 0.02 1.94 0.01 1.34 0.07 1.69 -0.05 2.09 -0.04 1.51

Agricultural and Biological Engineering

Precipitation: in-depth statistic analysis

Precipitation, a challenge in forecasting.
1 0.95

Semilog CDF -station #36

Florida climate controls: Latitude, land/sea distribution, ocean currents (Gulf Stream), migration of pressure systems and ENSO According to OBrien et al 1999: ENSO amplifies weather events in Spring, Summer and Fall, more frequent fronts; rainfall is above the average and temperature below normal. La Nia: fronts pass farther to the North, dryer conditions. Florida receives about 50 in of rainfall annually associated with frontal activity in winter (North) and afternoon storms in warm season (MayAugust). Agricultural and Biological Engineering
probability

0.9

0.85

0.8

0.75

0.7

0.65

0.6

x Observations * Model D1 o Model D2 o Model D3

0.55

0.5 -3 10

10

-2

10

-1

10

10

precipitation (in)

We combine hindcast climate modeling, bias correction and statistical analysis to extract variability components and generate 1-3 mos predictions. Not simple !

Model performance at different spatial resolutions

53 Stations Model & Observational Data Analysis: 1986-2006 Septembers
0.9 28.6

RMSE D1
0.2
0.2

28.6

RMSE D2
28.4

0.8 0.7

28.6

RMSE D3
28.4

Daily 28.4 precip. (in).

28.2 LAT.

0.2

0.2

0.2
0.6 0.5

0.2

LAT.

0.2

0.6

0.2

28

28

27.8

6 0.4 0. 0.2

0.4

0.6

0.40.2

27.8

0.2

0.4

0.4

0.3 0.2 0.1 27.6 27.8

0.4 0.8

0.6

0.4

0.4

LAT.

0.2

0.2
0.2
0.2
28.2

0.2

28.2

0.2

0.2
28

0.2

0.6

.8 00.6

0.2

0.6 0.4

0.4

27.6 -82.9 -82.8 -82.7 -82.6 -82.5 -82.4 -82.3 -82.2 -82.1 -82 LON.

27.6 -82.8 -82.6 -82.4 LON. -82.2 -82

-82.8

-82.6

-82.4 LON.

-82.2

-82

10 28.6

Monthly precip. 28.4 (in)

4
28.2

RMSE D1
6

28.6

RMSE D2
6
4
4 4

9 8

28.6

RMSED3

9
8

28.4

28.4

6
82 4

7 6 5 4 28.2 LAT.

5
4

LAT.

6
6

LAT.

4
27.8

28

27.8

3 2 1

27.8

8 7
6

27.6 -82.9 -82.8 -82.7 -82.6 -82.5 -82.4 -82.3 -82.2 -82.1 -82 LON.

27.6 -82.8 -82.6 -82.4 LON. -82.2 -82

27.6 -82.8 -82.6 -82.4 LON. -82.2 -82

7
4

28

28

5
5

4
2

6 8 0 2

28.2

7 5 4 6 9
6

1986-2006 September 53 stations

(unit: in) Obs. D1 Raw sim. D2 D3 D1 Bias corr. D2 D3

Mean daily Precip. for each station

mean 0.37 0.25 0.25 0.34 0.36 0.35 0.35 std 0.05 0.04 0.05 0.07 0.08 0.07 0.07 max 0.48 0.39 0.38 0.52 0.60 0.55 0.51 min 0.18 0.14 0.12 0.20 0.11 0.09 0.07

Mean Monthly Precip. for each station

mean 6.70 4.09 3.89 5.05 6.45 6.13 5.81 std 0.93 0.89 0.92 1.15 1.33 1.36 1.32 max 8.33 7.06 6.75 7.86 9.60 9.55 9.10 min 2.48 1.92 1.28 2.53 1.50 1.24 0.91

(unit: in) D1 Raw sim. D2 D3 D1 Bias corr. D2 D3

RMSE (daily total precipitation)

mean 0.25 0.25 0.25 0.30 0.28 0.24 std 0.18 0.17 0.18 0.21 0.19 0.17 max 0.85 0.78 0.84 0.94 0.80 0.86 min 0.13 0.13 0.13 0.16 0.15 0.12

RMSE (monthly total precipitation)

mean 5.04 5.13 5.99 4.78 4.54 5.29 std 1.25 1.26 1.33 1.48 1.38 1.27 max 9.33 8.49 9.43 9.48 9.03 8.74 min 2.46 1.86 3.03 2.02 2.11 2.24

Raw sim.: raw simulated results Bias corr.: bias corrected results D1, D2, and D3 represent the results for each domain 1, 2, and 3 respectively.

Points to Highlight, future work

We are developing a tool applicable in decision-making processes for water management. This tool will integrate a dynamical regional climate model, observational data and advanced statistical techniques to generate 13 month climate forecasting. Simultaneous with this on going project: Update LU distribution maps to improve climate hindcast simulations and reduce uncertainties in precipitation predictions, This investigation also provide a methodology to evaluate historical impacts of LUC on Floridas climate conditions. THANKS!!