Towards ecohydrological drought monitoring and prediction using a land data assimilation system: A case study on the Horn of Africa drought (2010–2011)

Despite the importance of the ecological and agricultural aspects of severe droughts, no drought monitoring and prediction framework based on a land data assimilation system (LDAS) has been developed to monitor and predict vegetation dynamics in the middle of droughts. In this study, we applied a LDAS that can simulate surface soil moisture, root‐zone soil moisture, and vegetation dynamics to the Horn of Africa drought in 2010–2011 caused by the precipitation deficit in two consecutive rainy seasons. We successfully simulated the ecohydrological drought quantified by the model‐estimated soil moistures and leaf area index (LAI). The root‐zone soil moisture and LAI are good indicators of prolonged droughts because they reflect the long‐term effects of past precipitation deficit. The precipitation deficit in 2010 significantly affected the land surface condition of the next rainy season in 2011, which indicated the importance of obtaining accurate initial soil moisture and LAI values for prediction of multiseasonal droughts. In addition, the general circulation model‐based seasonal meteorological prediction showed good performance in predicting land surface conditions of the Horn of Africa drought. Key Points We applied a land data assimilation system to the Horn of Africa drought in 2010–2011 We successfully monitor and predict both hydrological and ecological deficits due to drought The initial conditions of root‐zone soil moisture and LAI are important for drought prediction