Seasonal and Interannual Variability of Land-Atmosphere Coupling within the Southern Great Plains

This study investigated the temporal and spatial distributions of land-atmosphere coupling within the Southern Great Plains using the National Centers for Environmental Protection North American Regional Reanalysis (NCEP NARR) dataset from 1979 to 2014. Following the analysis of Wei and Dirmeyer (2012) which expanded land-atmosphere coupling into two segments including (1) the impact of soil moisture on surface fluxes (SM-E) and (2) the impact of surface fluxes on precipitation (E-P) this study examined land-atmosphere coupling via the effects of soil moisture variability on latent heat flux (SM-E) and the effects of latent heat flux variability on precipitation (E-P). To analyze the relationships, the correlation coefficient (r) and an index calculation were utilized. The index represented the product of the anomalies between the two variables, normalized by their respective standard deviations, and averaged over the number of days in a month. A positive index indicated that both variables had positive anomalies simultaneously or negative anomalies simultaneously. The larger the positive index, the larger the consistent anomalies were for both variables. Daily correlation analyses determined seasonal hot spots while monthly, grid analyses quantified the spatial variability within the domain. The results revealed a seasonal cycle for the domain averaged SM-E relationship, with an increasing linear relationship from the winter months to the summer months and a weakening linear relationship from the summer months and winter months. The correlation coefficient was found to be approximately .5 during the winter months and .9 during the summer months. The domain averaged E-P relationship revealed similar results, with a seasonal cycle of an increasing linear relationship between the winter months and summer months and weakening linear relationship between the summer months and winter months. Monthly averaged r values for 10-40 cm volumetric soil moisture and latent heat flux revealed a northeastward progression of increasing r values from January to June, with r values close to .9 throughout the domain from June to October, and a southwestward progression of decreasing r values from October to December. Monthly averaged r values for latent heat flux and twenty days of subsequent rainfall revealed similar results with increasing r values progressing northeastward from January to June, with the highest r correlations extending throughout the domain from June to August, then a southwestward progression of decreasing r values from August to December. An index calculation for latent heat flux and twenty days of subsequent rainfall in July from 1979 to 2014 revealed strong variability from one year to the next. An average monthly index calculation for the months of June, July, and August in drought, pluvial, and neutral precipitation periods during the June-July-August interval from 1979 to 2014 revealed high positive indices during drought months throughout the domain, with June having the highest indices. Neutral months exhibited both positive and negative indices within the domain, with index values close to 0. July was the only pluvial month with positive indices throughout the domain.