Search

Abstract: Sustainable decisions in water resources management require scientifically sound information on water availability. Central support in hydrological decision making arises from hydrological modelling which in turn depends on meteorological input. This work intends to show how integration of atmospheric science and hydrology, particularly the joint modelling of atmospheric and terrestrial water cycle, allows to provide decision support for two fundamental problems in sustainable water management: (1) the impact of global climate change on water availability, and (2) the near-real time estimation of recent resources and fluxes. The performance of joint atmospheric-hydrological simulations and its application for decision support is demonstrated for the Volta Basin of West Africa. First, the impact of global warming on water availability in the Volta Basin is assessed by joint regional climate-hydrology simulations. Time slices of ECHAM4 global climate scenario IS92a are dynamically downscaled to a resolution of 9×9km2 with the mesoscale meteorological model MM5 and the terrestrial water balance is subsequently calculated by the distributed hydrological model WaSiM. Results indicate a slight increase in total annual precipitation by 5%, but also a significant decrease of precipitation in April (in total 20%, for specific regions up to 70%), at the beginning of the rainy season. Terrestrial water balance variables react highly nonlinear to changes in precipitation and temperature. Second, the performance of joint atmospheric-hydrological simulations as a tool for near-real-time, model based water monitoring system for the White Volta subcatchment is assessed. Operational available GFS (Global Forecasting System) global analyses are dynamically downscaled to 9×9km2 resolution and meteorological fields subsequently applied for calculating the terrestrial water balance. Albeit a bias in precipitation simulations can be observed when comparing to point measurements, modelled river discharge agrees well with observations in 2004. It is concluded that particularly in areas with weak infrastructures, where few meteorological observations are available (or only in long delay), atmospheric models can provide reasonable meteorological information for hydrological modelling and therefore facilitate decision support. [Copyright &y& Elsevier]

Abstract: Savannahs are highly dynamic ecosystems but many of their properties and the related balances of energy, carbon, nitrogen, and water are still poorly understood. A particular scientific issue is the quantification of trace gases emitted from the soil of savannah ecosystems and their interaction with regional and global climate and air chemistry. Therefore it is important to develop and evaluate land-surface models that on the one hand represent vegetation and soil dynamics and on the other hand provide energy and water fluxes in a temporal resolution suitable for the application in climate/air chemistry models. In this paper, we present a consistent coupling between a common land-surface model (OSU) and a widely used biogeochemical model (DNDC) that is a first step for a full coupling of climate/air chemistry and biogeochemical processes. For consistency reasons, both models are linked to a general physiologically based plant model to provide the physical boundary conditions as well as the carbon and nitrogen in- and output variables. Evaluation is carried out with measurements of soil temperature, latent heat flux, soil water content, and soil emission data from two vegetation periods collected at a natural grassland site in Bontioli Nature Reserve, Burkina Faso (Africa). The results demonstrate that simulations of biogeochemical processes based on soil environmental conditions, calculated either with the land-surface model or with the unchanged biogeochemical model, do not differ significantly from each other. The OSU model simulates more realistic day-to-day variation of soil temperature as DNDC but the sensitivity of the biogeochemical simulation to this variation is small. In contrast, the sensitivity to differences in soil water content is high, but simulation results of both models are very similar on the daily scale and hardly depend on spatial soil resolution. [Copyright &y& Elsevier]