The effect of drought and landuse dynamics on water levels in lake Chilwa, Malawi.

Abstract

Climate change and land use change are the most important factors that determine the hydrological processes of the catchment, Drought occurrences triggered by climate change. coupled with inappropriate changes in land use are known causes of the lake water level fluctuations through changes in precipitation, evaporation and streamflow processes. The study investigated the effect of drought and land use/ land cover dynamics on Lake Chilwa water level fluctuations in Malawi during the years 1989 to 2016. It used a quasi-experimental design and quantitative research approach. Rainfall and temperature data from six stations of Chanco, Chikweo, Naminjiwa, Ntaja, Makoka and Zomba RTC and lake water levels data from four stations of Kachulu, Mposa, Namanja and Swang’oma in the basin were used for trend analysis. Furthermore, 12-month SPI time Series data were derived from rainfall using DrinC Software. Land use change was assessed using the extracted catchment shapefile in QGIS 2.18 with Trends Earth plugin from 2001-2015. NDVI time-series data were downloaded from Famine Early Warning Systems Network Climate Engine for land cover trends. Hydro-meteorological trends were tested using the Mann-Kendall trend test. The effect of drought and land use/land cover trends on lake water levels was assessed using a two-way ANOVA test. The quantitative land cover change detection between 2001 and 2015 revealed that cropland, grassland, water bodies and other lands declined by 1.71%, 2.36%, 7.95% and 5.32% respectively; while tree-coveted areas, artificial areas and wetland expanded. by 14, 15%, 84.41% and 19.10% respectively. The SDG 15.3. 1, land degradation indicator for improved, stable and, degraded land was 6.08%, 74.20% and 17.84% correspondingly NDVI trend increased insignificantly with a. maximum of 0.6782 in 2015/16 and minimum of 0.5493 in 1993/94. Annual mean lake levels trend decreased significantly with MK (-0.354), Sen’s slope (-0.035) and p-value (0.010 <0.05); while annual mean temperature trend increased significantly with MK (0.57), Sen's slope (0.032) and p-value (0.0001<0.05). However, the annual mean rainfall trend decreased insignificantly with MK (-0.103), Sen's slope (-3.256) and p-value (0.466>0.05), Lake Chilwa basin had experienced twelve drought years between 1989-2016. Two-way ANOVA results revealed no interaction effect between drought and land cover trends on lake water levels as the p-value (0.6584>0.05) and R-squared of 0.1720. Therefore, the Standardized Precipitation Index and Normalized Difference Vegetation Index are not good indicators of the Lake Chilwa water level fluctuations. Keywords: Drought, Land use dynamics, Lake water levels, Lake Chilwa basin