Global, 3-hourly, 9-km resolution soil moisture estimates are available with a mean latency of ~2.5 days from the NASA Soil Moisture Active Passive (SMAP) mission Level-4 Soil Moisture (L4 SM) product. These estimates are based on the assimilation of SMAP radiometer brightness temperature (Tb) observations into the NASA Catchment land surface model using a spatially distributed ensemble Kalman filter. Routine monitoring of the L4 SM system's assimilation diagnostics revealed occasionally large observation-minus-forecast Tb differences across eastern central Australia that resulted in large analysis increments (or adjustments) of the model forecast soil moisture. Because this region lacks in situ soil moisture measurements, we developed an alternative approach to assess the veracity of the soil moisture analysis increments in the L4 SM system. Using regional gauge-based precipitation data, we demonstrate that the L4 SM soil moisture increments are correlated with errors in the L4 SM precipitation forcing, suggesting that the SMAP Tb observations contribute valuable information to the L4 SM soil moisture estimates.