A SIMPLE SOIL MOISTURE INDEX TO FORECAST CROP YIELDS |
| |
| Authors: | Scott A Isard Mark R Welford Steven E Hollinger |
| |
| Institution: | 1. Department of Geography 607 S. Mathews , University of Illinois , Urbana, Illinois 61801;2. Department of Geology/Geography , Georgia Southern University , Stateboro, Georgia 30460;3. Office of Applied Climatology , Illinois State Water Survey , Champaign, Illinois 61821 |
| |
| Abstract: | Neutron probe soil moisture measurements obtained biweekly during the growing season between 1982 and 1991 from multiple depths under grass-covered plots at 17 Illinois Climate Network sites are used to forecast crop yields. A Soil Moisture Index (SMIX) that combines the effect of intensity, duration, and timing of drought or excessively wet conditions was computed by integrating the quantity of available soil moisture throughout the rooting zone over the growing season. Relationships between the SMIX values and crop yields are evaluated at county, regional, and statewide scales. Coefficients of determination (r 2) for relationships between the SMIX values and maize, soybeans, and hay yields at the statewide level are 0.88, 0.74, and 0.81, respectively, when the period of integration is terminated at the end of the growing season. This new soil index can be employed to forecast yields as early as 12 weeks before harvest for the state of Illinois. However, predictions with RMSE ≤ 10% of the mean yield can be achieved only for SMIX integration periods ending 5, 9, and 6 weeks before harvest for maize, soybeans, and hay, respectively. Nomograms are presented for using the relationships between the SMIX values and crop yields to forecast Illinois's major crops well before harvest. |
| |
| Keywords: | Phenology decision tree pioneer plants spectroscopy vegetation indices classification |
|
|
