Effect of Climate Change On Water Demand in Agriculture

You might also like

Download as pdf or txt
Download as pdf or txt
You are on page 1of 5

____________________________________________________________________________________________________

AWRA 2011 SPRING SPECIALTY CONFERENCE


Baltimore, MD
April 18-20, 2011 Copyright © 2011 AWRA
____________________________________________________________________________________________________

EFFECT OF CLIMATE CHANGE ON WATER DEMAND IN AGRICULTURE

Valerie Seidel, Christopher de Bodisco, Paul Yacobellis

ABSTRACT

Water supply constraints increasingly affect the choices that farmers make on crop selection, irrigation technology, or
even whether to remain in the industry. The uncertainty added by a changing climate makes these choices more difficult for
farmers and for water management agencies. The purpose of this paper is to estimate the effects of climate change on
agricultural water demand by considering farmers’ water use decisions across regions.
KEY WORDS: Agricultural Water Use, Climate Change in Agriculture, Irrigation, Crop Mix

INTRODUCTION

Water supply constraints increasingly affect the choices that farmers make on crop selection, irrigation technology, or
even whether to remain in the industry. The uncertainty added by a changing climate makes these choices more difficult for
farmers and for water management agencies.
The economics literature has addressed this issue from a number of perspectives. Adams et. al. (1995) estimate that a
doubling of CO2 in the climate will increase the economic returns from U.S. agriculture under GISS (Goddard Institute for
Space Studies) and GFDL-QFlux (Geophysical Fluid Dynamics Laboratory-Ocean heat flux version) climate forecasts, but
reduce economic returns by more than $25 billion under the more severe UKMO (United Kingdom Meteorological Office)
scenario. Much of the economic benefit arises from higher atmospheric levels of CO2, which are expected to increase crop
yields. Wolfe et al. (2007) makes the point that weeds are even more likely to benefit from higher levels of CO2. Schenkler
(2009) on the other hand, finds severe damage to U.S. grain crops if temperatures exceed 29 to 32 degrees C, with yields
increasing up to those threshold temperatures. Under the slowest warming scenario, yields are expected to fall over 30
percent by the end of the century. Hatch et al. (1999) finds increased yields in Georgia crops due to changes in precipitation,
CO2, and temperature resulting from climate warming. The study forecasts changes in crop mix, crop intensity, and potential
irrigation increases as non-irrigated land is irrigated in response to changing yields. Mimi (2010) estimates the water demand
effects of climate change on local farmers in the Middle East, finding that precipitation changes are critically important to
forecasted effects. Many studies emphasize the key role of changes in rainfall patterns.
All of the studies discussed above consider climate effects via the biological impact on crop yields and water needs. This
study takes a different approach. The demand for water is estimated across field crops, vegetable crops, and orchard crops
for all (multi-county) crop reporting districts in the 20 largest irrigating states. The study encompasses four censuses of
agriculture covering fifteen years. The purpose of this design is to estimate the impact of future changes in climatic


Respectively, Seidel, President; de Bodisco, Senior Economist; Yacobellis, Research Economist; all, of The Balmoral
Group, 341 North Maitland Avenue, Suite 100, Maitland, FL 32751, Tel. (407) 629-2185, Fax (407) 629-2183,
vseidel@balmoralgroup.us, cdebodisco@balmoralgroup.us.
1
2011 AWRA Spring Specialty Conference April 18-20, 2011

temperature and precipitation on water demand by measuring how current regional differences in those factors affect farmers’
water use decisions. To be effective, this approach requires that we control for other factors that affect the demand for water
in agriculture such as crop mix, soil quality, irrigation technique, water source, and prices. The process results in a more
consistent, detailed analysis of the effects of temperature and other geographic factors in addition to economic factors that
affect water demand.
Most water demand modeling treats water supply or availability as either completely fixed or instantly variable. We
propose that water is a quasi-fixed variable, i.e. one that changes, but changes slowly over time. Consequently, the amount of
water applied to any crop under irrigation can be varied (within permitted amounts) during the year, but extension of
irrigation to additional acreage requires investment in irrigation equipment or changes in permits that take time to complete.
This allows the estimated value of water to be greater or lower than its pumping cost in a given year, depending on supply
constraints, drought, or other factors. In addition, water use is measured over total harvested acres and not irrigated acreage.
The purpose is to emphasize farmers’ use of water (rather than irrigated acreage), and the choice to grow both irrigated and
rain-fed crops if that maximizes profit. Consistent with this approach, we define water use as acre feet of water applied per
harvested acre, rather than per irrigated acre as is more usual in the literature. This change can have large effects on perceived
water use. For example, the USGS (U.S. Geological Survey) reports that Arkansas applied an average of 1.97 acre-feet of
water per irrigated acre in 2000, whereas we report that Arkansas applied 0.53 acre-feet of water per harvested acre on
average over the study period. The USGS is reporting the intensity of water use on irrigated land, while we report the profit-
maximizing water use rate over all harvested farmland.
We posit that the value of water is determined by changes in farmers’ profit arising from farmers’ decisions to apply a
specific amount of water to grow a particular mix of crops. We estimate the change in profit as water use changes, while
holding all other factors constant. Thus the water value is not the cost of pumping water, but the estimated value of how
much profit increases when the only change is adding one more acre-foot of water. Once water values are estimated, the
impact of other factors on water value can be estimated by changing them one at a time.

Regional Variation in Water Demand

The estimated water demand function shows that the quantity of water demanded and the value of water vary
considerably across regions. On average, farmers apply three-quarters of an acre foot of water on each harvested acre. When
looking at districts that are the primary sources of high value crops such as vegetables and orchard crops, mean water use per
acre doubles to 1.5 acre feet per acre. These crop reporting districts include all those in states west of the Rocky Mountains,
Florida and coastal Texas. This region, the “vegetable region,” produces all types of crops, but supply the majority of high
value crops, producing over 80 percent of all the vegetables produced in the nation. When looking at districts growing
primarily field crops or grains, which include districts in the Midwest and South (the “grain region”), the opposite is true.
Mean water use is two-thirds lower, averaging 0.26 acre-feet per acre.
Profit per acre follows a similar pattern. When comparing the profit per acre across states, the states that apply the most
water per acre have much higher profit levels. While there a many factors that can account for these differences (crop mix,
input costs, soil quality and so on) the role of water is clear. By applying water until its value is essentially zero, farmers
2011 AWRA Spring Specialty Conference April 18-20, 2011

appear to be extracting all the return possible from water, resulting in much higher profits relative to other states. Thus
farmers are applying six times as much water per acre, and are reaping 2.5 times more profit per acre.

RESULTS: EFFECTS OF VARIATION IN MEAN ANNUAL TEMPERATURE

While it may be less intuitive to discuss water demand in terms of the value of water rather than the quantity of water, it
can provide additional insights, particularly concerning climate warming. The water demand model estimates that every ten
percent increase in mean annual temperature raises the mean value of water by 30 percent. Each one degree increase in
temperature increases the value of water by approximately $14 per acre foot. In other words, holding all other factors
constant, an increase in temperature increases the value of water, regardless of how much water is being used. If mean annual
temperatures increase, the overall demand for water in agriculture will shift out, increasing both the quantity of water
demanded and the value of water. Because of the wide range of annual temperatures among the states in this study, variation
in temperature account for as much as 37 percent of the variation seen in water values across states.
The same pattern is not observed when comparing temperature and the quantity of water demanded directly. Overall,
there is relatively little correlation between mean annual temperature and the quantity of water demanded. Some states with
high mean temperatures use very little water (Georgia) and some use a great deal of water (Arizona). This should not be
interpreted to say that annual warming would have no effect on the quantity of water demanded in agriculture. Rather it
shows that the mechanism by which temperature affects water demand is through the value of that water. That is, the
important role temperature plays on the demand for water is not apparent until other factors that affect water demand, such as
crop mix decisions, input costs, and so on, are controlled for in the analysis. Thus we should say that with other relevant
factors held constant, climate warming will increase the demand for water by increasing the value of water in agriculture.
One indication that this mechanism exists is that the states where water values are most sensitive to temperature are the
states that apply the most water per harvested acre – all of which are states west of the Rocky Mountains. Among these states,
variation in water quantity explains as much as 80 percent of the dollar effect on water values caused by temperature
differences. That is, as the quantity of water applied per acre increases, the impact of temperature on water values increases
as well. For example, in Arizona, which on average applies over 4 acre-feet of water per harvested acre, a mean annual
temperature of 20.5 increases water values by $189, while in Idaho, which applies an average of 1.3 acre-feet of water per
harvested acre, a mean annual temperature of 8.45 increases water values only $78.

IMPLICATIONS

Many studies of the effects of climate change on specific crops have forecast increased benefits from irrigation, and
consequently higher water demand. We find a similar result by estimating the effect of temperature differences over time and
across regions. On average, a ten percent increase in mean annual temperature increases water values by 30%. As the value of
water increases in agriculture, the quantity of water demanded will also.
This analysis suggests two very different forecasts of the effects of climate change on water use. In both cases
increasing mean annual temperatures increases the value of water in agriculture. In one case farmers respond to increasing
water values by increasing irrigation incrementally. Currently we are seeing incremental changes in water use. The share of
2011 AWRA Spring Specialty Conference April 18-20, 2011

vegetables that are irrigated has remained relatively constant, at just under 90 percent, but the share of field crops that are
irrigated is growing at approximately 2 percent per year. Most analyses of the effect of climate change on specific crops
forecast a similar scenario.
In the second case, farmers, particularly in the Southeast, adopt water use patterns currently seen in the West and in
Florida. We have seen how farmers in the West grow more high-value crops, apply much more water per acre, and reap much
higher profit rates per acre. Under what conditions could farmers in the Southeast adopt a similar water use pattern? There are
several factors that suggest that such a change is possible. The Southeast has abundant forest and pasture land that is
available to be converted to crop production, and it has untapped sources of water yet available. In addition, major changes in
crop mix have occurred in the recent past in this region. Between 1949 and 1987 soybean production changed from a minor
forage crop to the second largest source of farm revenue, which resulted in a significant change in the types of soils under
tillage.
On the other hand, there are important obstacles that could prevent major changes in farmers’ behavior. The first obstacle
is farmers’ access to available water. Water rights in the Southeast are primarily governed by regulated riparian water law. In
the West, water access is governed by appropriative rights (“first in time, first in line”), and in Florida water access is entirely
dependent on a permit system with no property right to water. There are indications that riparian law itself does not prevent
water from being applied at high levels designed to maximize yields (see Frederick, 1996), but the combination of legal and
social factors that enable water to be used as intensively as it is in the West, may not exist in other regions going forward.
A second obstacle is competition from imports. Price competition from low-cost imports, among other factors, has kept
the real price of crops (even vegetables) flat or falling over the last two decades or longer. For fresh vegetables in particular,
the timing of harvests is a critical component of where, and how profitably, high-value crops are marketed. Climate change
will undoubtedly change when these windows appear, and which regions may see an opportunity develop.
Economics generally forecasts future behavior based on previous behavior, so it is not well-suited to predict structural
changes in behavior. What we can do, is show what conditions are necessary for major changes to occur. This paper has
identified several reasons why major regional changes in agricultural water demand are conceivable. Because there are
myriad other factors that should be considered, we hope future climate warming forecasts will expand crop mix models
beyond major crops that are currently being cultivated to include less-likely crops that could have larger impacts.

REFERENCES

Adams, Richard M., Ronald A.Fleming, Ching-Chang Chang, Bruce A.McCarl, Cynthia Rosenzweig. 1995. “A
reassessment of the economic effects of global climate change on U.S. agriculture.” Climate Change 30(2): 147-167.
Frederick, Kenneth D., Tim VandenBerg, and Jean Hanson. 1996. “Economic Values of Freshwater in the United States.”
Resources for the Future Discussion Paper 97-03.
Hatch, Upton, Shrikant Jagtap, Jim Jones, and Marshall Lamb. “Potential effects of climate change on agricultural water
use in the southeast U.S.” JAWRA Journal of the American Water Resources Association 35(6):1551-1561.
Mimi, Ziad A., and Sireen Abu Jamous. 2010. Climate change and agricultural water demand: Impacts and adaptations.”
African Journal of Environmental Science & Technology Vol. 4(4):183-191.
2011 AWRA Spring Specialty Conference April 18-20, 2011

Schlenker, Wolfam, and Roberts, Michael J. 2009. “Nonlinear temperature effects indicate severe damages to U.S. crop
yields under climate change.” PNAS September 15, 2009 vol. 106 no. 37 15594-15598

USDA/Census. 1997. “Farm Income Data.” Washington, DC. http://www.nass.usda.gov/census.

USDA/ERS. 2000. “Farm Income Data.” Washington, DC. http://www.ers.usda.gov/Farm Income/finfidmu.htm>.

USDA/NASS. 2000. “Crops County Data.” Washington, DC. http://www.usda.gov/nass/pubs/histdata.htm>.

USDA/NASS. 2000. “Historical Data.” Washington, DC. http://www.usda.gov/nass/pubs/histdata.htm>.

USDA/NASS. 2000. “Prices Received by Farmers.” Washington, DC. http://www.usda.gov/nass/pubs/histdata.htm>.

Wolfe, David W., Lewis Ziska, Curt Petzoldt, Abby Seaman, Larry Chase, and Katharine Hayhoe. 2008. “Projected change
in climate thresholds in the Northeastern U.S.: implications for crops, pests, livestock, and farmers.” Mitig Adapt Strat
Glob Change 13:555–575.

You might also like