Abstract:
Lake Rukwa, one of Tanzania’s most important endorheic lakes, has shrunk
dramatically over the past three decades due to climate variability,
sedimentation, and human-induced land use change. This study applied multi
temporal remote sensing using Landsat imagery from 1994, 2004, 2014, and
2024, supported by Sentinel-2 and Google Earth Pro validation. The
Normalized Difference Water Index (NDWI) was used to delineate lake surface
area, while the Normalized Difference Turbidity Index (NDTI) served as a proxy
for sedimentation at major river inflows. Land use and land cover (LULC)
changes were classified with a Random Forest algorithm.
Results reveal a net loss of about 65,000 hectares of lake surface area, with
the sharpest decline between 2004 and 2014 at –0.68 percent per year.
Extensive deforestation, estimated at 700,000 hectares, and cropland
expansion exceeding 500,000 hectares have intensified soil erosion and
sediment inflows. Rising NDTI values in rivers such as the Songwe (0.08 to
0.24) confirm worsening turbidity. These pressures have accelerated sediment
accumulation, reduced water depth, and destabilized the lake’s hydrological
balance.
The findings highlight human-driven catchment degradation as the dominant
driver of Lake Rukwa’s decline. Mitigation requires basin-level reforestation,
erosion control, and sustainable water abstraction, supported by integrated
monitoring and adaptive management strategies..
