Determination of drought probabilities in Texas and a comparison to agricultural impacts: a decile climatology

Abstract

There are a variety of drought indicators used to monitor drought conditions across the United States as well as in other countries. Some of these indicators often transform data sets to approximate known distributions in order to perform various statistical tests for the analysis of drought conditions. An attempt was made to assess the potential for drought conditions in Texas through the modification and use of existing drought indices, such as the Standardized Precipitation Index (SPI) and the deciles technique, in order to assist the agricultural community while, at the same time, preserving the integrity of the data. It was determined that for periods of record equal to or greater than 45 years, the structure of the monthly precipitation distribution stabilizes and does not change significantly over the years. In addition, the use of explicit time scales in drought monitoring is of great use since a distinction can be made between different durations of drought. Shorter-term droughts have a greater temporal fluctuation and occur in a more random pattern than the longer-term drought periods. An analysis of the 6-, 12-, and 24- month drought combinations indicated that more 24-month droughts occur alone than do the 6-, 12-, or shorter-term drought periods. In addition, the 6+12+24-month drought combination occurs more frequently in the western half of Texas without any prior indication from a l-, 2-, or 3-month drought. The western half of Texas is also prone to dry conditions in all months, with the exception of May. This was deduced by observing individual month plots of the ratio of the first decile values (lowest 10%) to the median values, which gives an indication of the structure of the lower half of the monthly precipitation distribution. In addition, precipitation occurring at or below the 5th decile range during a majority of the growing season tends to result in lower crop yields for both sorghum and wheat. A multiple month drought analysis revealed that several longer-term drought periods existed where individual month patterns were not discernible for sorghum. For wheat, the pattern was not nearly as clear.