Drought Information Needs for Water Resources Management: Texas as a Case Study
Bridget R. Scanlon, Rong Fu, Todd Caldwell, Di Long, and Nelun Fernando
Jackson School of Geosciences, Univ. of Texas at Austin
Outline
• Background on aquifers• Drought information needs (ET, SM, GW)
– Modeling and satellite products• Example: 2011 drought in Texas
Storativity: amount of water released from an aquifer for a decrease in water level.
Unconfined aquifers1 – 30% of water storage can be recovered
Confined aquifers 0.005 – 0.5% of water in storage can be recovered
Implications for Using Model and Satellite Data
• Ogallala Aquifer: demand driven system, not really impacted directly by changes in recharge but indirectly affected by pumpage
• Confined aquifers: disconnected from surface water, recharged thousands of years ago, only impacted by drought indirectly through pumpage
Information Needs
• Groundwater planning time horizon: 50 yr, 2010 – 2060• Based on 1950s drought• National drought maps and archives with time series are very
valuable, extend back to 1900?• Climate in next 50 yr: 1 yr extreme droughts? Longer term droughts? • How well can we predict beginning of droughts? Persistence?, end
of droughts? • Seasonal drought predictions: fall – early spring: water transfers if
drought predicted• Not irrigate if predict intense drought? • Daily temperature data for electricity sector
Data Needs
• Evapotranspiration (MODIS, Landsat, NLDAS): – Irrigation needs
• Soil moisture storage (SMAP, NLDAS): – Irrigation needs – Drought persistence…feedback– Precipitation requirements for reservoir recovery– Flood predictions
Data Needs
• Groundwater (NLDAS, GRACE)• Recharge (NLDAS): P − ET = Recharge 500 ± 50 − 480 ± 48 = 20 ± 98 • GRACE:∆TWS = ∆ SMS + ∆ GWSWater Scarcity = Demand > Supply
Satellite and Modeling Products
• Satellite ET, GRACE derived ET, and LSM ET• Satellite ET unconstrained by water balance (1 km
resolution)• Model ET more reliable, coarse ( 14 km resolution)• GRACE GW, LSM GW• Difficult to disaggregate GRACE TWS into
components with large variability in LSM SMS• LSM provide upper BC for groundwater
(unconfined aquifers)
PDSI and GRACE Total Water Storage
-120
-80
-40
0
40
80
120
km
3
-200
-150
-100
-50
0
50
100
150
200
Totalw
aterstorageanomalies(m
m)
GRGS RL02CSR RL05PDSI
-10
-8
-6
-4
-2
0
2
4
6
8
10
PDSI
2003 2004 2005 2006 2007 2008 2009 2010 2011 2012
-40
0
40
80
120
(mm)
-40
-20
0
20
40
60
80
km
3
Precip. anomal
Long et al., 2013
Soil Moisture Storage
2003 2004 2005 2006 2007 2008 2009 2010 2011 2012
-80
-40
0
40
80
120
km3
-150
-100
-50
0
50
100
150
200
Soilmoistureandtotalw
aterstorage
anomalies(mm)
NLDAS-2 Noah
NLDAS-2 Mosaic
GLDAS-1 Noah
GLDAS-1 Mosaic
GLDAS-1 VIC
GLDAS-1 CLM
CSR RL05
GRACE Gridded Product
2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013
Wa
ter
sto
rage
ano
ma
lies
(mm
)
-200
-150
-100
-50
0
50
100
150
200
(km3)
-120
-90
-60
-30
0
30
60
90
120
CSRGFZJPLAlgorithm1
2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013
Precipanomaly(m
m)
-100-50
050
100150
-50
0
50
100
(km3)
Summary
• Information needs for drought:– Long-term planning, 1 yr versus multiyear droughts?– Seasonal forecasting: 6 – 9 months– Daily data: temperature, heat waves– Archive data back to 1900
• Satellite data: ET, irrigation demand; soil moisture
• Modeling: upper boundary condition for unconfined aquifers