The nature of the relationship between soil-moisture and subsequent precipitation is debated. Past studies have indicated that it varies by location, but there have been disagreements on feedback sign, strength and statistical significance over the same locations even while using the same data sets due to differences in statistical approaches. Here we re-evaluate estimates of feedback strength using Granger causality (following Tuttle and Salvucci 20016) for SMOS and AMSR-2soil moisture products and find that there is evidence both of there being a positive feedback over the entire nation (SMOS) and of there being a west-to-east gradient of positive to negative values (AMSR-2, similar to what was found by Tuttle). This project has demonstrated that differences arise using identical statistical estimation techniques with different satellite estimates of soil moisture. The passive microwave AMSR-E and AMSR-2 data produce a gradient of positive to negative impact factors across the U.S. while the SMOS data set yields a mostly positive impact factor over the entirety of the U.S. Here we explain this discrepancy by noting that the areas where the sign of the impact differ are areas where the soil moisture perturbations of SMOS and AMSR-2 are anticorrelated, and also that these areas coincide with areas of high vegetation cover, which are well known to challenge remote sensing estimates of moisture.

To determine hydrologic changes in a warmer climate, we impose precipitation and potential evaporation perturbations on hydrologic response functions constructed from precipitation and satellite soil moisture observations across the United States. Despite nonlinearities in the evaporation (E) and drainage (D) responses and opposing-sign perturbations, changes in individual fluxes are superposable. Empirical frameworks (Budyko) can misrepresent changes in E/D partitioning by neglecting shifts/trends in hydrologic regime and subseasonal precipitation dynamics. E/D both increase to balance mean precipitation (P) increases, and increased E reduces soil moisture. E and D are generally more elastic to changes in P than E. The results suggest that (1) the impacts of regional hydrologic perturbations may allow for simple superposition/scaling, (2) changes in timing/intensity of precipitation may have substantial impacts on mean moisture states and fluxes, and (3) changes to the distribution of surface moisture states are likely more relevant for E/D partitioning than common aridity indices.

We present a kernel auto-regressive (KA) method which can be used to represent the daily to multi-day auto-correlation structure of precipitation time series, using information both in the occurrence and intensity of measured rainfall events. The method is able to capture a larger fraction of the memory in multiple time series than commonly-used occurrence-based Markov chain models, even when the intensity distribution is allowed to be conditioned on the Markov state. The KA method is less sensitive to the spatial scale at which the data is reported, as it is not strictly reliant on patterns of wet and dry days for providing correlation. Output from the KA model can be used as weather generator model simulations, as empirical representations of process structure, as representation of weather/climate variability partitioning, or as climatological null models against which observations can be compared for statistical significance. The KA method demonstrates improvements in each of these over classic occurrence Markov chain models and daily independent climatology, in both representations of interannual precipitation variability and in downstream water balance variables. We provide climate null confidence intervals for precipitation trends (driven largely by autumn increases), and decompose variability into trend, interannual, and weather components (in increasing order of magnitude) for the Contiguous United States.