Effects of Cloud Seeding on Wheat Production in Oklahoma

Whether or not cloud seeding actually enhances precipitation, some conclusions may still be drawn regarding changes in wheat production that would occur as a result of increased rainfall.  On average, increased rainfall after 1 March can be shown to result in higher wheat yields.  By using the CERES-wheat crop model, some potential benefits of cloud seeding can be quantified.

In this analysis, wheat production was modeled for two locations: Canadian County and Texas County, in Oklahoma.  Recorded wheat yields from the Oklahoma Department of Agriculture were compared with modeled wheat yields for a multi-year period to determine whether the CERES-wheat model is a good estimator at each site.  Once this determination was made, the model was run for several rainfall enhancement scenarios using the same multi-year weather data from each site.  The mean percentage difference in yields predicted for baseline and enhanced precipitation data sets was used as a measure of tangible benefit to Oklahoma wheat productivity.

Canadian County


The National Weather Service maintains first order, daily records of parameters such as high and low temperatures, precipitation and solar radiation at the Will Rogers Airport in Oklahoma City.  The long record length and high quality of this data set make it attractive for crop model validation.  Because Canadian County is the closest major wheat producing county to the Will Rogers Airport observing station, wheat yields predicted using Oklahoma City weather data were used as estimators for Canadian County wheat yields.

Chart 1 is a comparison of  predicted and recorded wheat yields for Canadian County on a standardized scale.  The close correspondence between modeled and measured wheat yields indicates that the model is good estimator of final yield performance for this particular location and data source.

Chart 2 shows changes in modeled wheat yields induced by randomly enhanced precipitation events after 1 March of each weather record year.  For each harvest year, the model indicates that increases in precipitation of up to 20% after March 1 result in increased wheat yields.  Chart 3 shows modeled percent yield changes as a function of various rainfall enhancement scenarios.

To estimate the of economic benefit of increased yields, the total yearly harvests for each county are multiplied by the predicted percent yield change for the specific year in question.  The average of the resulting bushel increases is then multiplied by an approximate price per bushel.  When these calculations are performed for Canadian County, the average increase in revenue due to 10% enhancements in half of all precipitation events after 1 March is found to be about $72,000.

Texas County


A combination of National Weather Service Cooperative station records and Oklahoma Mesonet data was used for the Texas County analysis.  Cooperative records from Goodwell and Hooker, Oklahoma were combined to create a continuous record of high and low temperatures and precipitation from 1961 through 1996.  Daily solar radiation totals from the Goodwell mesonet site were used to model solar radiation for the Cooperative station records, which lack any radiation data.

Chart 4 compares  predicted and recorded wheat yields for Texas County on a standardized scale.  The poor correlation between modeled and measured wheat yields indicates that the model provides a less reliable estimator of final yield performance for this particular location and data source than it does for the Canadian County location and data source.  Consequently, economic impact estimates based on modeled Texas County wheat yields are less certain than those based on Canadian County modeled wheat yields.

Chart 5 shows changes in modeled wheat yields induced by randomly enhanced precipitation events after 1 March of each weather record year.  Much like the Canadian County results, the model indicates that enhanced precipitation events in Texas County after 1 March generally result in higher yields.  Chart 6 shows modeled  percent yield changes as a function of various rainfall enhancement scenarios.

The model indicates that a 10% increase in half of all rain events between 1 March and harvest in Texas County would, on average, increase wheat production by nearly 129,000 bushels.  This increase would, in the absence of price adjustments, result in a revenue increase of approximately $387,000.

Comparisons and Conclusions


While projected wheat revenue benefits for Texas County resulting from artificially enhanced rainfall far exceed those for Canadian County, it should be stressed that they are also much more uncertain.  This fact is born out by the poor correlation between modeled and measured wheat yields for Texas County.  Some of the gap in reliability between projections for Texas County and Canadian County stems from the different weather data sources used for each location.  Two incomplete records are combined in the Texas County analysis and augmented with simulated solar radiation data while a single record containing measured solar radiation data is used in the Canadian County analysis.

Beyond the factors mentioned above, only very general soil data was used for both locations, leaving open the possibility that the soil type used for either location could be unrepresentative of the soil types in which most of the counties' wheat is grown.  In addition, only one planting date, 15 September, was assumed for both sites.  An average of yields, weighted by the distribution of acreage planted with respect to date, is needed to correctly account for the effect of planting date on final yield performance.

An assumtion has also been made in this study that the CERES-wheat model treats artificial, random increases in rainfall with the same accuracy as it treats year-to-year fluctuations.  As yet, there is no way to validate this assumption.

It can be said with certainty that a 10% increase in half of all rainfall events between 1 March and harvest would, on average, produce tangible benefits in the form of increased wheat production for the counties studied in this analysis.  It should be emphasized that the practice of cloud seeding in Oklahoma has not yet been proven to produce such increases in precipitation.
 



last updated 8/27/98