Tropical Lake Levels and Their Relationship to Rainfall
Abstract
The availability of satellite altimeters and improvements in satellite estimates of river and lake levels are offering an exciting monitoring alternative to currently limited prediction systems using current climate models. Aware of existing limitations in data retrievals, we have developed a simple linear model for estimating lake level as a function of freshwater flux into the catchment basin for 12 tropical lakes and reservoirs: 8 in Africa, 3 in Central and South America, and 1 in Southeast Asia. In our model three parameters, effective catchment basin, time delay, and drainage timescale are determined from linear regression based on the simultaneous availability of remotely sensed lake level and rainfall. We present results of estimates of net surface freshwater flux and lake levels during a 16-year period (1992-2007). Comparison between two different altimeter satellite-based lake level datasets shows very good agreement for most lakes. For net freshwater flux (i.e., rainfall minus evaporation), we use three different rainfall products: the European Centre for Medium Range Weather Forecasts (ECMWF) ERA-Interim reanalysis, the Global Precipitation Climatology Project (GPCP) rainfall, and the Tropical Rainfall Measurement Mission (TRMM) 3B42 precipitation index rainfall. ERA-Interim evaporation is combined with each of the three rainfall products to form three estimates of net surface freshwater flux. Results from models are denominated as Model-I, Model-G, Model-T, respectively. A comparison of rainfall products shows differences, and as a result the best model for a given lake varies. The median correlation between the observed LEGOS and Model-G lake levels is significantly higher than for Model-I, with the median RMS difference between observation and model slightly lower for Model-G than for Model-I. For many tropical lakes the best results are obtained using one of the observation-based products, GPCP or TRMM. All three model results show that all lakes, except Lake Turkana, have pronounced seasonal cycles, with the largest seasonal cycles occurring for the lakes in high rainfall regions: Kainji in Africa, Balbina in South America, and Lake Tonle Sap in Southeast Asia. With the seasonal cycle removed, the east African rift valley lakes (Turkana, Tanganyika, Mweru) show pronounced rises in 1997-8 in response to the combined effects of El Niño and the Indian Ocean dipole. In contrast, the Central and South American lakes (Nicaragua and Balbina) show significant level decrease for the same time period. There is also evidence of rapid increases in lake level in response to tropical cyclones (Nicaragua and Malawi). At short intraseasonal periods, Lake Tonle Sap shows evidence of a 30-60 day fluctuation driven by rainfall fluctuations partly associated with the Madden-Julian Oscillation. These results show noticeable climate impacts on tropical lake levels.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2009
- Bibcode:
- 2009AGUFM.H11C0812R
- Keywords:
-
- 1804 HYDROLOGY / Catchment;
- 1807 HYDROLOGY / Climate impacts;
- 1874 HYDROLOGY / Ungaged basins;
- 1876 HYDROLOGY / Water budgets