Climatic anomalies pose severe challenges for decision makers and resource managers to mount effective responses in a timely manner. This is a particularly significant problem with respect to gradually developing anomalies such as droughts. The goal of this project is to determine the importance of climatic forecasts to individual, state, and private sector planning and management through a retrospective assessment of drought responses in Oklahoma in 1996. Specifically, the impact of the drought on the wheat crop, and the possibility that predictive information might have reduced some of the losses, is examined.
This is accomplished through a combined modeling approach using data from the Oklahoma Mesonet, a point-process precipitation model, and the CERES-Wheat Model operating under the Decision Support System for Agrotechnology Transfer (DSSAT). The DSSAT model takes into account an extensive range of soil, climatic, and plant physiological variables, as well as cropping patterns to predict the yield for the wheat crop. This model was run by incorporating climatic information for 1995-6, along with a range of information based on the historical climate of the area. Thus, the model was first run using actual data from 1995-6 through Jan. 1, and then was run repeatedly using the type of climatic conditions that might be found from Jan 1 to June 1. The DSSAT model was run using data from 1995-96 until Jan. 7, and a series of runs using the range of climatic possibilities from Jan. 7 to June 1, etc. The results show the range of potential outcomes, as represented by a series of box plots. The results also illustrate the point at which all possible climatic outcomes were predicting a significantly low wheat yield. Based on anecdotal evidence of the 1995-6 drought, which suggested that farmers who planted at different times experiences different yields, the model was run based on different planting dates. Results indicate that there is indeed a noticeable difference in the modeled wheat yields given the different planting dates.
Combined with a point-process precipitation model, this approach produces information regarding the likelihood of extreme precipitation events and their impact of crop yield. The information provided by this approach can provide a powerful tool to decision makers during periods of drought or other climatic extremes.