Browsing by Author "kavubu oscar Godfrey"

Now showing 1 - 2 of 2
  • No Thumbnail Available
    Item
    USING NDVI AND NDWI INDICES TO ASSESS THE SPATIAL, TEMPORAL AND SEASONAL SPREAD OF AQUATIC WEEDS ON LAKE WAMALA, UGANDA (2017–2025).
    (2026-08-18) kavubu oscar Godfrey
    Aquatic weed infestation poses a growing threat to the ecological integrity and socio-economic value of shallow tropical lakes in East Africa. Lake Wamala, a shallow lake in central Uganda, supports important fisheries and provides water for surrounding communities, yet its weed dynamics remain poorly quantified. This study assessed the spatial distribution, temporal trends and seasonal dynamics of aquatic weed cover on Lake Wamala using the Normalized Difference Vegetation Index (NDVI) and the Normalized Difference Water Index (NDWI) derived from Sentinel-2 satellite imagery. A dataset of 30,296 pixel-level observations was extracted from 16 ground stations distributed across five regions and seven ecological zone types, spanning three years (2017, 2022 and 2025) and five months (January, February, March, June and September). Descriptive statistics, weed-cover classification using NDVI thresholds, correlation analysis and spatial–temporal trend analysis were performed. Group differences were tested by Welch’s and one-way analysis of variance, and the NDVI–NDWI association by Pearson correlation. The overall mean NDVI was −0.073 (SD = 0.153) and the overall mean NDWI was 0.206 (SD = 0.169), indicating that open water dominates the lake surface. Weed-cover classification showed that 95.2% of pixels were open water or low vegetation, 1.3% moderate vegetation and 3.5% high vegetation (dense weeds). The Southern region emerged as the principal weed hotspot, with the only positive regional mean NDVI (0.012) and the highest proportion of dense-vegetation pixels (12.02%), followed by the Eastern region (4.93%). NDVI declined sharply between 2017 (−0.037) and 2022 (−0.113) before partially recovering by 2025 (−0.072), a pattern attributed to the 2020–2023 La Niña-driven high water levels in the Lake Victoria basin. Seasonally, January (dry season) recorded the least negative NDVI (−0.011) while February (onset of the rains) recorded the most negative (−0.133). NDVI and NDWI were strongly and inversely correlated overall (Pearson r = −0.971), with zone-level coefficients ranging from −0.743 to −0.989, confirming their value as complementary indices for weed mapping. All four group-difference tests — by zone type, region, year and month — were significant at p < .001, with season producing the largest effect (η² = 0.08) and year the smallest (η² = 0.02). The study concludes that aquatic weed infestation on Lake Wamala is not a uniform, lake-wide problem but a spatially concentrated one, driven by nutrient enrichment, shallow bathymetry and sheltered conditions in the Southern and Eastern regions, and modulated by strong inter-annual climate variability. NDVI and NDWI from freely available Sentinel-2 imagery provide a reliable, low-cost basis for routine monitoring. It is recommended that management be spatially targeted on the Southern and Eastern regions, embedded within an adaptive framework that responds to climate variability, supported by an integrated satellite-plus-ground monitoring system, and complemented by catchment nutrient management and community engagement. These measures would allow limited management resources to be directed where they yield the greatest ecological and economic return. Keywords: Lake Wamala; aquatic weeds; NDVI; NDWI; Sentinel-2; remote sensing; weed mapping; eutrophication; adaptive management.
  • No Thumbnail Available
    Item
    USING NDVI AND NDWI INDICES TO ASSESS THE SPATIAL, TEMPORAL AND SEASONAL SPREAD OF AQUATIC WEEDS ON LAKE WAMALA, UGANDA (2017–2025).
    (Kavubu oscar Godfrey, 2026-08-18) kavubu oscar Godfrey
    ABSTRACT Aquatic weed infestation poses a growing threat to the ecological integrity and socio-economic value of shallow tropical lakes in East Africa. Lake Wamala, a shallow lake in central Uganda, supports important fisheries and provides water for surrounding communities, yet its weed dynamics remain poorly quantified. This study assessed the spatial distribution, temporal trends and seasonal dynamics of aquatic weed cover on Lake Wamala using the Normalized Difference Vegetation Index (NDVI) and the Normalized Difference Water Index (NDWI) derived from Sentinel-2 satellite imagery. A dataset of 30,296 pixel-level observations was extracted from 16 ground stations distributed across five regions and seven ecological zone types, spanning three years (2017, 2022 and 2025) and five months (January, February, March, June and September). Descriptive statistics, weed-cover classification using NDVI thresholds, correlation analysis and spatial–temporal trend analysis were performed. Group differences were tested by Welch’s and one-way analysis of variance, and the NDVI–NDWI association by Pearson correlation. The overall mean NDVI was −0.073 (SD = 0.153) and the overall mean NDWI was 0.206 (SD = 0.169), indicating that open water dominates the lake surface. Weed-cover classification showed that 95.2% of pixels were open water or low vegetation, 1.3% moderate vegetation and 3.5% high vegetation (dense weeds). The Southern region emerged as the principal weed hotspot, with the only positive regional mean NDVI (0.012) and the highest proportion of dense-vegetation pixels (12.02%), followed by the Eastern region (4.93%). NDVI declined sharply between 2017 (−0.037) and 2022 (−0.113) before partially recovering by 2025 (−0.072), a pattern attributed to the 2020–2023 La Niña-driven high water levels in the Lake Victoria basin. Seasonally, January (dry season) recorded the least negative NDVI (−0.011) while February (onset of the rains) recorded the most negative (−0.133). NDVI and NDWI were strongly and inversely correlated overall (Pearson r = −0.971), with zone-level coefficients ranging from −0.743 to −0.989, confirming their value as complementary indices for weed mapping. All four group-difference tests — by zone type, region, year and month — were significant at p < .001, with season producing the largest effect (η² = 0.08) and year the smallest (η² = 0.02). The study concludes that aquatic weed infestation on Lake Wamala is not a uniform, lake-wide problem but a spatially concentrated one, driven by nutrient enrichment, shallow bathymetry and sheltered conditions in the Southern and Eastern regions, and modulated by strong inter-annual climate variability. NDVI and NDWI from freely available Sentinel-2 imagery provide a reliable, low-cost basis for routine monitoring. It is recommended that management be spatially targeted on the Southern and Eastern regions, embedded within an adaptive framework that responds to climate variability, supported by an integrated satellite-plus-ground monitoring system, and complemented by catchment nutrient management and community engagement. These measures would allow limited management resources to be directed where they yield the greatest ecological and economic return. Keywords: Lake Wamala; aquatic weeds; NDVI; NDWI; Sentinel-2; remote sensing; weed mapping; eutrophication; adaptive management.