Data from TRMM have demonstrated great value in applications ranging from water budget studies to operational hurricane forecasting. However, direct use of the data in operational hydrologic forecasting has been challenging due to the relatively infrequent refresh rate of the data relative to the 1-6 h periods used in hydrologic forecasting. In this work, TRMM data have been combined with other data from geostationary and low-Earth orbit satellites by the Self-Calibrating Multivariate Precipitation Retrieval (SCaMPR) algorithm to produce quantitative precipitation estimates (QPEs) that are available at high temporal frequency and thus well-suited for operational hydrologic forecasting.

Validation experiments over the entire CONUS domain suggest that the primary benefit of TRMM ingest is bias reduction, though other skill measures also show improvement at lower latitudes. Comparisons with gauge-only QPEs for 22 catchments in Texas and Louisiana indicated that the SCaMPR analysis outperformed the gauge-only QPEs in comparisons with operational gauge-radar basin average precipitation estimates, and that ingest of TRMM further improved the detection of heavy rain by SCaMPR. Hydrologic simulations using the National Weather Service operational lumped model calibrated using the operational gauge-radar estimates indicated that adding TRMM to SCaMPR also improved simulation of annual runoff; however, for the majority of the basins the gauge-only QPEs still outperformed SCaMPR in simulating flood events.