The Global Hydro-Intelligence Subseasonal-to-Seasonal (GHI-S2S) Forecast System

Producing reliable and timely global hydrometeorological forecasts on subseasonal-to-seasonal timescales is a significant need of many government agencies, companies and operating centers, especially in relation to disseminating near real-time drought and flood potential information and making life-saving decisions. To help meet this need, the Global Hydro-Intelligence Subseasonal-to-Seasonal (GHI-S2S) system has been developed at NASA, which derives its skill from: (i) accurate initial conditions produced by an offline land modeling system through the assimilation of various satellite data; and (ii) an ensemble of meteorological forecast data produced by state-of-the-art ocean-land-atmosphere forecast systems. GHI-S2S uses NASA’s Land Information System (LIS) land modeling and hydrological framework, which employs a variety of ensemble and forecast capabilities along with improved initial land surface or hydrologic conditions for the forecasts, including assimilation of remotely sensed soil moisture data. GHI-S2S also utilizes precipitation hindcasts and forecasts from the North American Multi-Model Ensemble (NMME), and non-precipitation data (e.g., near-surface temperatures and winds) from NOAA’s Climate Forecast System, version 2 (CFSv2). GHI-S2S is configured for the full global domain (including Antarctica), producing daily and monthly hydrometeorological forecast fields out to nine lead months. A suite of output metrics and products for drought and flood potential prediction are generated both for hindcasts and routine forecast runs. We present real-world examples and verification of the system for different domains and hydrometeorological variables (e.g., soil moisture, terrestrial water storage and streamflow).