Development of a Soil Moisture Index for Agricultural Drought Monitoring using a Hydrologic Model SW

1
Development of a Soil Moisture Index for
Agricultural Drought Monitoring using a
Hydrologic Model (SWAT), GIS and remote sensing
By
R. Srinivasan Director and Associate
Professor Spatial Sciences Laboratory Texas AM
University

• Balaji Narasimhan
• Graduate Research Assistant
• Dept. of Agricultural and Biological Engineering
• Texas AM University

2
Outline

• Problem statement
• Palmer Drought Severity Index (PDSI)
• Limitations
• Can we do better?
• Soil Moisture Index
• Results/Discussion
• Conclusion

3
Problem Statement

• USA annual loss 6-8 billion (FEMA, 1995)
• Texas 1998 drought - 5.8 billion
• Drought is a normal recurrent feature of climate
• Drought cannot be prevented
• However through proper monitoring the losses can
  be minimized

4
Palmer Drought Severity Index (PDSI)

• PDSI is widely used for agricultural drought
  monitoring
• Uses a simple water balance model to estimate ET,
  runoff, recharge
• From historical average then estimates the
  potential precipitation needed to meet these
  hydrological demands

• Precipitation deficit Drought Index

5
Limitations of PDSI

• Simple lumped parameter water balance model
• uniform soil/land-use properties across the
  entire climatic zone (7000 to 100,000 km2)
• Thornthwaite method to estimate ET
• No effect of land-use on runoff!!
• One index for the entire climatic zone

6
Can we do better??

• Advances in GIS and remote sensing
• Availability of spatially accurate soil,
  land-use/land-cover and climatic data
• Improved spatially distributed hydrologic models
• A better representation of real-world
• Increased computational power

7
Objective

• To develop a drought index based on soil moisture
  deficit by incorporating GIS and remote sensing
  technology and the distributed parameter
  hydrologic model SWAT (Soil and Water Assessment
  Tool)

8
Methodology

• Soil moisture deficit Drought Index
• Crop growth depends on soil moisture

• Need a hydrologic model to estimate soil moisture
  content
• Soil and Water Assessment Tool (SWAT)
• Widely tested and accepted
• Can be applied locations in snow cover conditions
• Applied for entire USA (HUMUS)
• Integrated with GIS

9
Study Area Trinity River Basin
10
Methodology

• Input data
• Daily precipitation, Max. and Min. Temperatures
  (91 stations, from 1901-1998).
• STATSGO soil data
• MRLC Land Use/Land Cover data
• Resolution - 4km X 4km for integration with
  NEXRAD
• Upper Trinity 1854 sub-basins
• Lower Trinity 950 sub-basins
• Model Calibration and Validation
• Using measured USGS stream flow data
• 8 locations in Upper Trinity and
• 4 locations in Lower Trinity

11
Model Validation (Upper Trinity)
R2 0.73
12
Model Validation (Lower Trinity)
R2 0.70
13
Soil Moisture Deficit Ratio

• The model was run using weather data from 1960
  to 1997
• The soil moisture deficit ratio is then
  calculated as

14
(No Transcript)
15
(No Transcript)
16
Soil Moisture Index

• Soil moisture deficit ratio
• dryness during a given month.
• The cumulative soil moisture deficit ratio
• rate at which the dryness/drought is progressing.
  
• Cumulative soil moisture deficit ratio for all
  the sub-basins were analyzed to determine the
  worst drought from 1 to 24 months
• A procedure very similar to the one adopted by
  Palmer is used to categorize different drought
  classes

17
(No Transcript)
18
Soil Moisture Index

• The Soil Moisture Index during a given month is
  calculated by

19
(No Transcript)
20
Garza-Little Elm
Grapevine
21
(No Transcript)
22
County average SMI (Upper Trinity)
23
County Average SMI (Lower Trinity)
24
Conclusion

• Preliminary analysis show that SMI compare well
  with PDSI.
• Improved spatial resolution (16km2)
• Incorporating NEXRAD will further enhance the SMI
  estimates
• Detailed analysis will be conducted using weather
  data from 1901 till date

25
Conclusion

• Stochastic nature of drought will be incorporated
  in the index
• Time step will be reduced from monthly to weekly
• Time series models will be developed for drought
  forecasting

26
http//webgis.tamu.edu
27
(No Transcript)
28
(No Transcript)
29
(No Transcript)
30
(No Transcript)
31
(No Transcript)