Northern Prairie Wildlife Research Center
The presence of a negative correlation of the variations of precipitation and temperature in many climate regions (Karl and Quayle, 1981; Madden and Williams, 1981) may have masked the temperature effect on PDSI by precipitation effect. Additionally, fluctuations of precipitation with large amplitudes in some regions strengthens the precipitation effect on PDSI. The PDSI in these situations represents the effect of precipitation anomalies on soil water conditions. However, as we showed, even in these climate regions temperature effects contribute to the PDSI. Our calculations showed that temperature effect dominated the sign of PDSI when precipitation anomalies were relatively small. For regions that show an absence of a negative correlation of variations of temperature and precipitation, a statistically significant negative correlation between PDSI and temperature variations indicated the impact of temperature on the PDSI.
These results have been obtained using data from two regions in the central United States that have contrasting humid and dry temperate climate conditions. Although regional limitations may prevent a direct generalization of some of the results to other climate regions in the world, this study sheds light on the importance of temperature effect on the PDSI.
The temperature effect on the PDSI complicates the interpretation of precipitation anomalies using the PDSI and, thus, application of the index to inferring precipitation variations, particularly from reconstructed PDSI. One major uncertainty is whether the reconstructed PDSI from tree-rings can be interpreted as an indicator of past precipitation variations. Because of this uncertainty, the historical wet and dry periods attributed to reconstructed PDSI variations may not be caused by precipitation anomalies alone. Therefore, we suggest caution in applying the PDSI in climate studies as well as climate monitoring.