Characterization of soil moisture variability over South America: linkages to remote sources of variability

A. CELESTE SAULO; CAROLINA VERA; PABLO C. SPENNEMANN

Qingdao

Conferencia; CLIVAR Open Science Conference; 2016

Slowly evolving variables, like sea surface temperature (SST) and soil states (e.g. soil moisture), have the potential to increase the atmospheric predictability through the influence on the surface fluxes, affecting thereby the boundary layer development and finally the atmospheric circulation at different time scales. During the last 30 years the influence of local and remote SST conditions over the climate in South America has been extensively studied. However, the influence of the land surface conditions on the predictability of that particular region has not been deeply investigated yet. A better comprehension of land-atmosphere interactions requires an improvement in the understanding of all the components of the hydrological cycle from a regional perspective. A basic aspect of the hydrological cycle, like the main characteristics of the soil moisture (SM) variability is barely documented. A better knowledge of SM variability has highly relevance, because locations with high SM variability are related with regions of potentially strong land-atmosphere interactions. The absence of a detailed analysis of SM climatology over South America could be explained by the lack of observational data. To fill this gap, this work documents the mean fields and main modes of SM variability at different soil depths over South America, using 29 years (1980-2008) of monthly mean 1°x 1° data derived from the Global Land Data Assimilation System (GLDAS) version 2.0. In addition the linkages with sea surface temperature (SST) are analyzed. Over Southern South America, part of the La Plata Basin (LPB), a local maximum of SM anomalies variability is observed. For deeper layers (0-100 cm), this variability is mostly explained by the low frequencies (i.e internannual). Furthermore the interannual variance of the low-pass filtered SM anomalies exhibits a similar spatial distribution to that obtained for the SM anomalies leading pattern based on an EOF analysis. This confirms the relevance of the interannual variability in explaining the SM variance in LPB. Lagged correlation between SST (leading) and SM anomalies averaged over SESA show significant negative correlations over central equatorial Pacific from lags -30 to -12 months, which reaches its maximum negative value on lag -24. Furthermore, this pattern resembles very much that associated with precipitation variability in the same region, as extensively described in previous works. In special the SST anomaly pattern is highly associated with ENSO, which exhibits the largest influence on precipitation variability in SESA from austral spring of the year in which ENSO develops to austral fall of the next year. The results found in this study could have potential implications in forecast improvements of SM conditions over SESA at seasonal time scales. This could be achieved combining coupled models forecasts and statistical methodologies, using for instance the SST as a predictor of SM conditions. Still, further studies are needed to analyze the potential improvements in prediction of SM and its implications to the socio-economic sectors.