A Methodology for Evaluating Climate Models and Reanalysis Products: Use of Observed Stream Flow to Diagnose Atmospheric and Soil Water Balance
Abstract
Climate models are the primary tools for predicting future climate and for creating credible future climate change scenarios. However, before using these models to simulate future climate conditions, we need evaluate their skill in simulating the current and past climates. Here, we propose that for climate studies that aim to simulate changes in future stream flows over large basins such as the Nile or the Congo basins, the climate model used should be evaluated in its ability to simulate observed stream flow which is a key variable of the soil water balance. The long term average stream flow over a large basin is exactly balanced by the long term average convergence of atmospheric moisture, a key variable of the atmospheric water balance. Our analysis uses three types of inputs: climate model simulations results, reanalysis products, and observations of stream flow, rainfall and evaporation. Three different study areas with diverse climatic conditions and different degree of complexity of topographical conditions are considered: the Sahara desert, the Congo basin and the Nile basin. Simulations from the Hadley model with different spatial resolutions, and ERA-INTERIM reanalysis dataset are evaluated. Both the Hadley model and the ERA-INTERIM are capable of simulating the seasonal cycle of the hydrologic variables. However, the magnitudes of these variables are significantly biased compared to observations, even at the highest resolution. Using a moving average approach to reduce noise in the atmospheric moisture flux field is shown to enhance accuracy of the atmospheric water balance over the complex topography of the Nile basin. We demonstrate that the spatial resolution of the climate model should be evaluated in the context of the scale of the region of interest and the degree of complexity of surface topography.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2011
- Bibcode:
- 2011AGUFMGC11B0923S
- Keywords:
-
- 1833 HYDROLOGY / Hydroclimatology;
- 3337 ATMOSPHERIC PROCESSES / Global climate models