Projet de recherche S0/00/024 (Action de recherche S0)
Context and objectives
This project is the continuation of a former project, CLIMLAKE that started in 2001, extending until the beginning of 2005 and also financed by the SPO (EV/13/02). This project addressed the recent concerns about climate variability (global warming, ENSO,...) and its effect on ecological processes in Lake Tanganyika. The main components of the project were environmental monitoring, study of present plankton, geochemical studies, diatoms analysis in sediment cores and development of an eco-hydrodynamic model. In this project, the potential contribution of Remote Sensing (RS) data was underestimated. This is why the CLIMFISH project was launched. The general objectives of CLIMFISH are (1) to investigate the impact of recent environmental changes on the fisheries of Lake Tanganyika using weather and limnological monitoring, fisheries statistics, remote sensing and eco-hydrodynamics modelling; (2) to use those results and the ECO-Hydro model, developed in the former research program, under expected climate change most likely scenarios to evaluate the future of the fisheries at Lake Tanganyika; (3) to evaluate the feasibility of developing an ECO-FISH model as a forecasting tool for the managers of Lake Tanganyika fisheries.
Methodology
1. Field monitoring (weather, limnology, fish statistics and biology) from 11-2004 to 04-2006 (continuation of a 3 years CLIMLAKE monitoring).
2. Development of a remote sensing methodology for the determination of lake surface T° and Chl-a.
3. Refinement and validation of the ECO-hydro model using RS and monitoring data.
4. Comparison of the model outputs with fish statistics for the recent period (2002-06).
5. Use of historical data (environment and fisheries) to check model adequacy.
6. Use of model approach for preliminary scenarios useful for fisheries predictions.
Concerning the RS aspects of the research, the “water case” has been analyzed. The different MODIS product levels and several correction methods to extract water leaving radiance and Chl-a concentration have been tested using the cruises and permanent stations data. Daily cloud free images have been chosen, downloaded and batch-processed to produce a 3.5 years data base.
Results
Relationships between surface chlorophyll a and chlorophyll a in the euphotic zone have been established. Primary production data (48 measurements for the period 2002-04) showed a range for the period 2002-05 between 382-1653 mg C m-2 day-1 while average was 675 mg C m-2 day-1. RS data download (MODIS) is finished for the period 2002-05. The period untill 12-2006 is presently downloaded. Test of several algorithms of atmospheric correction and Chlor-a concentration computation have been performed. Batch processing are finished for the first period and in course for the last year. RS Missing data filling has been applied to the first period data set. It should be applied again to the whole period from 07-2002 untill 12-2006. The validation using the in situ data from the permanent stations of Kigoma and Mpulungu has been performed. RS outputs of the first period have been exchanged with other teams RS output interdisciplinary (fisheries-limnology and modelling) interpretation is in course. Fish biology and statistics have been compiled for the 3 main species of the lake. Relationships were found with RS data (transparency). Four type of fish catch variation were identified. Patterns of variability are checked for predictability investigation. Primary productivity of the Lake under actual circulation and solar radiation using the ECO-hydro model was compared with regional weather patterns to study the climatic influences on lake ecosystem since the year 1970. The simulated primary production is being analysed with fish catch data.
Products and services
RS products :
1. Temporal series of low resolution images (1 km) providing Chlor-a concentration, SST and K490..
2. Tables showing the results of the “training” phase.
3. Tables showing the results of the validation.
4. Tables of the spatial generalization process at the hydrodynamical model resolution.
5. Spatio-temporal regionalization of the lake.
