Abstract
LDAS-Monde is a global land data assimilation system (LDAS) developed by Centre National de Recherches Météorologiques (CNRM) to monitor land surface variables (LSV) at various scales, from regional to global. With LDAS-Monde, it is possible to jointly assimilate satellite-derived observations of surface soil moisture (SSM) and leaf area index (LAI) into the interactions between soil biosphere and atmosphere (ISBA) land surface model (LSM) in order to analyze the soil moisture profile together with vegetation biomass. In this study, we investigate LDAS-Monde’s ability to predict LSV states up to two weeks in the future using atmospheric forecasts. In particular, the impact of the initialization, and the evolution of the forecasted variables in the LSM are addressed. LDAS-Monde is an offline system normally driven by atmospheric reanalysis, but in this study is forced by atmospheric forecasts from the European Centre for Medium-Range Weather Forecasts (ECMWF) for the 2017–2018 period over the contiguous United States (CONUS) at a 0.2° × 0.2° spatial resolution. These LSV forecasts are initialized either by the model alone (LDAS-Monde open-loop, without assimilation) or by the analysis (assimilation of SSM and LAI). These two forecasts are then evaluated using satellite-derived observations of SSM and LAI, evapotranspiration (ET) estimates, as well as in situ measurements of soil moisture from the U.S. Climate Reference Network (USCRN). Results indicate that for the three evaluation variables (SSM, LAI, and ET), LDAS-Monde provides reasonably accurate and consistent predictions two weeks in advance. Additionally, the initial conditions after assimilation are shown to make a positive impact with respect to LAI and ET. This impact persists in time for these two vegetation-related variables. Many model variables, such as SSM, root zone soil moisture (RZSM), LAI, ET, and drainage, remain relatively consistent as the forecast lead time increases, while runoff is highly variable.
Highlights
Extreme meteorological and climatic events, such as heatwaves and droughts, are predicted to increase in frequency and magnitude in future decades [1,2]
While land surface models (LSM) simulations provide temporally and spatially continuous information about land surface variables (LSV), they are by no means perfect, and often lack complex interactions and physics that result in predictions differing from reality
The objectives of this study are to assess to what extent (1) LSV conditions can be forecasted using an LSM, (2) LSV initial conditions influence the forecasts, (3) data assimilation can improve the accuracy of initial conditions of LSV forecasts, and (4) LSV forecasts can benefit to crop monitoring
Summary
Extreme meteorological and climatic events, such as heatwaves and droughts, are predicted to increase in frequency and magnitude in future decades [1,2]. The monitoring, prediction, and warning of droughts, floods, famine, and other extreme events can be accomplished with land surface models (LSM) [10,11] that can simulate the LSV responses to these extreme events. These events and their responses pose a significant scientific challenge for the adaptation to climate change [1]. Good knowledge of both land and lower atmosphere conditions is necessary to accurately monitor and predict LSV values. While LSM simulations provide temporally and spatially continuous information about LSV, they are by no means perfect, and often lack complex interactions and physics that result in predictions differing from reality
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
