In this paper, we review the use of satellite-based remote sensing in combination with in situ data to inform Earth surface modelling. This involves verification and optimization methods that can handle both random and systematic errors and result in effective model improvement for both surface monitoring and prediction applications. The reasons for diverse remote sensing data and products include (i) their complementary areal and temporal coverage, (ii) their diverse and covariant information content, and (iii) their ability to complement in situ observations, which are often sparse and only locally representative. To improve our understanding of the complex behavior of the Earth system at the surface and sub-surface, we need large volumes of data from high-resolution modelling and remote sensing, since the Earth surface exhibits a high degree of heterogeneity and discontinuities in space and time. The spatial and temporal variability of the biosphere, hydrosphere, cryosphere and anthroposphere calls for an increased use of Earth observation (EO) data attaining volumes previously considered prohibitive. We review data availability and discuss recent examples where satellite remote sensing is used to infer observable surface quantities directly or indirectly, with particular emphasis on key parameters necessary for weather and climate prediction. Coordinated high-resolution remote-sensing and modelling/assimilation capabilities for the Earth surface are required to support an international application-focused effort.

Weather extremes have harmful impacts on communities around Lake Victoria in East Africa. Every year, intense nighttime thunderstorms cause numerous boating accidents on the lake, resulting in thousands of deaths among fishermen. Operational storm warning systems are therefore crucial. Here we complement ongoing early warning efforts based on numerical weather prediction, by presenting a new satellite data-driven storm prediction system, the prototype Lake Victoria Intense storm Early Warning System (VIEWS). VIEWS derives predictability from the correlation between afternoon land storm activity and nighttime storm intensity on Lake Victoria, and relies on logistic regression techniques to forecast extreme thunderstorms from satellite observations. Evaluation of the statistical model reveals that predictive power is high and independent of the type of input dataset. We then optimise the configuration and show that false alarms also contain valuable information. Our results suggest that regression-based models that are motivated through process understanding have the potential to reduce the vulnerability of local fishing communities around Lake Victoria. The experimental prediction system is publicly available under the MIT licence at http://github.com/wthiery/VIEWS.

Integration of simulated and observed states through data assimilation as well as model evaluation requires a realistic representation of soil moisture in land surface models (LSMs). However, soil moisture in LSMs is sensitive to a range of uncertain input parameters, and intermodel differences in parameter values are often large. Here, the effect of soil parameters on soil moisture and evapotranspiration are investigated by using parameters from three different LSMs participating in the European Land Data Assimilation System (ELDAS) project. To prevent compensating effects from other than soil parameters, the effects are evaluated within a common framework of parsimonious stochastic soil moisture models. First, soil parameters are shown to affect soil moisture more strongly than the average evapotranspiration. In arid climates, the effect of soil parameters is on the variance rather than the mean, and the intermodel flux differences are smallest. Soil parameters from the ELDAS LSMs differ strongly, most notably in the available moisture content between the wilting point and the critical moisture content, which differ by a factor of 3. The ELDAS parameters can lead to differences in mean volumetric soil moisture as high as 0.10 and an average evapotranspiration of 10%-20% for the investigated parameter range. The parsimonious framework presented here can be used to investigate first-order parameter sensitivities under a range of climate conditions without using full LSM simulations. The results are consistent with many other studies using different LSMs under a more limited range of possible forcing conditions.