Relative contributions of initial hydrologic conditions and seasonal climate forecast skill to seasonal hydrologic prediction globally

Global seasonal hydrologic prediction is crucial to mitigating the impact of flood and drought related losses, especially in the third world. The potential for a system capable of making such forecasts has been elusive, but now appears feasible with the advent of global hydrologic models. Hydrologic prediction skill at seasonal lead times (i.e. 1-6 months) comes from the knowledge of initial hydrologic conditions (IHCs - primarily the state of initial soil moisture and snow) and seasonal climate forecast (CF) skill. Understanding the relative contributions of these factors to seasonal hydrologic prediction skill at different forecast initiation dates and lead times is important for identifying those regions where useful skill can be attained in any given season, given current global hydrologic monitoring capability (the basis for providing the IHCs) and seasonal CF skill. In this work we seek to quantify the contributions of IHCs and CF skill to global seasonal hydrologic prediction skill on a relative basis across different seasons. We report the results of two model-based experiments (1) Ensemble Streamflow Prediction (ESP) and (2) Reverse-ESP over a 33 year reforecast period (1971-2003). We use the Variable Infiltration Capacity (VIC) model, which has been extensively used at global scale, for both experiments. We compare cumulative runoff (CR) and soil moisture (SM) forecasts obtained from each experiment with a control simulation forced with observed atmospheric forcings over the reforecast period and estimated Root mean square error (RMSE) of both experiments for each forecast initialization date and lead time. Our initial results indicate that there are regions (mainly snowmelt dominated) where the IHCs dominate the seasonal hydrologic prediction skill for as ,much as 6 months. On the other hand in some regions CF skill dominates throughout most of the year at all lead times.