Effect of lake area change by basin development on the rainfall at the environs of Tonle Sap Lake in Cambodia

Tonle Sap Lake is known for its unique hydrological characteristics. Its water surface at the end of rainy season is around 5 times as large as that at the end of dry season. This seasonal change is caused by the seasonal change of the water level of Mekong River which connects to Tonle Sap Lake through Tonle Sap River. During dry season, lake water is drained to Mekong River and lake area gets shrinking. On the other hand, during rainy season, a reverse flow occurs at Tonle Sap River and river water of the Mekong flows into the lake, which makes the lake larger and larger.

However, this annual cycle may change in future. Recently, lots of dams are under planning and under construction in the upper Mekong River basin. Since dams will regulate the seasonal variation of water level, water level difference between the Mekong and the lake may become smaller and then the less water of the Mekong flows into the lake. As a result, seasonal expansion of Tonle Sap Lake may become smaller.

Considering this possibility, we carried out a sensitivity study of the lake area to the rainfall at the environs of Tonle Sap Lake using a regional climate model as a kind of an environmental impact assessment.

The Advanced Regional Predication System (ARPS) coupled with the Simple Biosphere Model 2 (SiB2) is adopted for this study together with in-situ observations at several stations over and around the lake. It is configured with three nested grids centered in the middle of the lake: the first grid covers 2,100km x 2,100km domain including whole Indochina Peninsula, the South China Sea, and the Andaman Sea with 30 km x 30km grid elements; the third grid, the target domain, is composed of 4 km x 4km grid elements covering 400 km x 400 km area. To account for synoptic conditions, the National Centers for Environmental Prediction (NCEP) final reanalysis (FNL) produced every 6 h are used as the initial and boundary conditions for the first nested grid. The initial surface conditions are also derived from the NCEP-FNL data except that the lake surface temperature is given by the in-situ observed value (30.6 degree) instead of the reanalysis value (25.5 degree). Since large scale atmospheric conditions has significant effects on the atmospheric field during the summer and winter monsoon seasons and local effects are not expected to be dominate, we excluded these seasons and targeted pre-monsoon and post-monsoon seasons.

We did three numerical simulations for each season: (i) control run which describes current conditions (CNTRL), (ii) a simulation with fixed lake area for both seasons (FIXED), and (iii) a simulation assuming all of the lake area is converted to the agricultural land (AGRIC). Lake area is 1,600km2 (pre-monsoon) and 12,304km2 (post-monsoon) for CNRTL runs, 3,728km2 for FIXED runs (for pre- and post- monsoon seasons), and 0km2 for AGRIC runs.

During the pre-monsoon season (at the end of April), when the lake disappears and is converted to agricultural land (AGRIC experiment), rainfall over the (former) lake increases and rainfall over the northwestern part of the (former) lake decreases. When the lake gets larger (FIXED experiment), rainfall over the northwestern part of the lake decreases, too. There are only small changes in rainfall at other parts surrounding the lake. It suggests that rainfall over the northwestern part is sensitive to the lake area and is largest at the current small lake condition: in both cases with larger lake and no lake, the rainfall there will decrease. From the analysis of MTSAT cloud top temperature, it has been shown that the northwestern part of the lake is the most active part of the convective cloud activity among the areas around the lake. It is considered to be attributed to the convergence of the lake breeze and sea breeze and thus rainfall will decrease if lake disappears.

During the post-monsoon season (at the end of October), it rains a lot at the western part of the lake at the CONTRL run. When the lake becomes smaller (FIXED experiment) or disappears (AGRIC experiment), this rain area gets smaller and weaker. Predominant rainfall at the western part of the lake has been realized by the ground rain gauges as well and thus this rainfall is considered to be brought by the current expanded lake.

In Cambodia, it has been known that it rains even in the pre-monsoon and post-monsoon seasons. This study suggests that rainfalls during these seasons are at least in part brought by Tonle Sap Lake and its current condition with small area in the pre-monsoon season and large area in the post-monsoon season. If seasonal changing pattern of the lake area changes due to the basin development, rainfall at the environs of the lake will change (decrease in many of the cases).