This proposal is for an �Integrated Science Data Analysis� project that will specifically address the following ESE strategic science questions: 1) How can weather forecast duration and reliability be improved by new space-based observations, data assimilation, and modeling? And; 2) how are variations in local weather, precipitation and water resources related to global climate variation? We propose to improve operational streamflow forecasts by assimilating remotely sensed snow cover products from the Moderate Resolution Imaging Spectroradiometer into the hydrologic models used by the National Weather Service. In such models, proper initialization (updating) of model states is a key component to accurate simulation of snowmelt and streamflow a range of time and space scales. Currently, remotely sensed snow cover and snow albedo data are not used to update hydrologic states within operational forecast models despite their inherent utility in snowmelt simulations and long historical record - one of the longest of any remotely sensed land-surface feature. Rather, daily updating of model states for many forecast points is conducted in an ad hoc manner by an experienced river forecaster based on their understanding of the forecast basin, models, and prediction errors. Inaccuracies in these subjective interpretations are exacerbated by climatic extremes as conceptually based forecast models may perform poorly during climatic conditions not represented in the historical record. Such conditions may be occurring with increased frequency as recent climate analyses have shown widespread declines in the winter snowpack in mountains of the western USA and Europe associated with positive temperature anomalies. Hence, there is some urgency in shifting operational forecasts toward techniques capable of exploiting advances in remote sensing and data assimilation techniques. In the context of this NRA, we propose to derive pixel-specific parameterizations of snow covered area and snow-surface albedo data from MODIS for use in the SNOW17-SACSMA model. Our aim is illustrate the utility of remotely sensed data within operational hydrologic models. In this regard, we will develop a system that allows forecasters to utilize remotely sensed data, adjust filter parameters, evaluate uncertainty, and assess the accuracy of the predicted update - such a system will build confidence in the forecasts because operational hydrologists will still be able to apply their expertise at all stages. Our approach will make the minimum number of adjustments to the current forecasting system needed to facilitate the inclusion of the remotely sensed data, ensuring transferability of our results to operations.
