Presented in Fulfilment of the Requirements for the Degree of Doctor of Philosophy in Water Resources Engineering and Management (Surface Water Management)
No Thumbnail Available
Date
2024-12
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Addis Ababa University
Abstract
In many places across the world, non-point source pollutants (NPSP) are the primary threat to aquatic ecology and the primary cause of surface water deterioration. Because pollutants and waste may be dumped easily into the surface waters, they are especially vulnerable to pollution. NPSs contaminate surface water more than point sources do. Fertilizers, herbicides, oil, silt, and nutrients from animals and malfunctioning septic tanks can all contribute to NPSP. It has several negative impacts on aquatic life, including their invasion and a rise in the amount of suspended matter that blocks light and damages plant life in water. Eutrophication is the outcome of an increase in the rate at which ecosystems receive organic materials. For example, Koka Reservoir regularly experiences algal blooms due to the reservoir's nutrient load. Remedial measures are complex due to their widespread nature. To assess the problem of water quality and the effects of water quality (WQ), a systematic review was conducted using PRISMA (Preferred Reporting Items for Systematic Review and MetaAnalysis Statement). The synthesis points out the following gaps: Inadequate evidence-based research on the effects of contaminated water on agriculture, health, and socioeconomics; a lack of combined spatial and temporal surface water quality data and monitoring; a lack of relative contribution from non-point sources; a lack of detailed integrated spatial and temporal water quality impact; and policy responses for surface water quality drivers for the case studies have received scant attention nationally and in the basin for the identified gaps. This research work is designed to evaluate the non-point source pollutants load, prioritize the subwatersheds based on the sediment load, LULC, and morphometric characteristics using hydrologic modeling, and develop a remote sensing-based water quality model. Several strategies were used to accomplish the designated goals. We used a variety of primary and secondary data sources. Analysis was done on the land use and land cover in 2023 and 2003. Primary data was gathered, including modeling work data and lake water quality analysis from the Koka Reservoir. For a year, water quality data for non-point source pollution modeling was collected every month from the Melka Kunture gauging stations. The remote sensing technology was used to determine the spatiotemporal water quality condition of the Koka reservoir and develop a water quality model. To study the spatiotemporal dynamics of water quality indicators at Koka Reservoir, a Sentinel-2 satellite remote sensing data collection was used to develop a water quality monitoring model. For the annual scale, the temporal studies of water quality parameters were carried out from 2017 to 2022; for the monthly scale, it was carried out from June 2021 to May 2022. Regression analysis and empirical model development were used to develop algorithms that correlated satellite reflectance data with in situ measurements of turbidity (TU), total suspended matter (TSS), and chlorophyll-a (Chl-a). The determination coefficients (R2 ) for each of the examined parameters are higher than 0.67, suggesting that they can all be used to develop predictive models. Good correlations between field-based and calculated Chl-a, TU, and TSS have been found, as indicated by the developed algorithms' respective R2 values of 0.91, 0.92, and 0.67.
The multi-criteria decision-making, AHP-VIKOR, and hydrological modelling were used for subwatershed prioritizing. The findings indicated that 2524.6 km2 (25.45%) is in the high sensitivity class, 2722.14 km2 (27.44%) is in the moderate sensitivity class, 854.35 km2 (8.61%) is in the low sensitivity class, 2205.48 km2 (22.23%) is in the very low sensitivity class, 1611.43 km2 (16.25%) is in the class of very high erosion sensitivity. We evaluated and created linear function models to estimate WQ indicators, including Chl-a, turbidity, and TSS, using the Sentinel-2 image band ratios of B5/B4, B4/B3, and B4, respectively. The built-up area and agricultural land grew by 80.15% and 147.29%, respectively. There was a reduction of 47.55%, 96.7%, and 74.37% in forest, grassland, and shrubland, respectively. The BASIN and SWAT models are the most effective for assessing point and NPS contamination in various basins. SWAT was used to examine the spatial and temporal variation of the non-point source pollutants NO3 í , PO4 í, TN, and TP. The 2003 and 2023 LULCs were the main data sources used to evaluate the change in NPSP loads. The Melka Kunture gaging station non-point source pollution modeling was calibrated and validated between 2009 and 2014 and 2015 and 2019, respectively. The sensitivity analysis led to the selection of nine nutrient-related parameters for calibration. The most critical parameters are the phosphorus uptake distribution (R_P_UPDIS.bsn), the phosphorus percolation coefficient (10 m3/Mg) (R_PPERCO.bsn), and the organic P settling rate (R_RS5.swq). For the calibration and the validation periods, the results revealed good and very good performance. While the mean annual increase in surface runoff ranges from 183.1 mm to 487.9 mm, the mean annual increase in sediment yield ranges from 25.46 to 27,298.75 t. Runoff ranged from a minimum of 10.69 mm (5.1%) to a maximum of 223.3 mm (66.5%). The PO4 í load went from 3.12 to 2459.7 kg, and the NO3 í burden went from 127.6 to 20,739.7 kg. The TP load went up from 1383.5 to 133,641.3 kg, and the TN load went up from 4465.5 to 482,014.5 kg. According to the monthly nitrate loading analysis, the “Belg” season, the second rainy time from February to May when rainfall is highly variable in time and space, has a higher nitrate load than the rainy season, probably due to nitrification. The LULC alteration increased surface runoff and NPSP loads (nitrate, phosphate, total nitrogen, total phosphorous, and sediment). The study demonstrated that Sentinel-2A-derived regression models can support the spatiotemporal estimation and mapping of the annual and monthly patterns of Chl-a, TU, and TSS over the Koka reservoir. This enables improved capacity to analyze reservoir status and strategies for water resource management. The algorithms could potentially be useful as a monitoring tool for water quality in other regions in the country or other data-scarce areas of the world with comparable environmental and hydro-climatic contexts. The low operational cost of using freely available remotely sensed imagery is a strong incentive for water agencies to complement their field campaigns and produce spatially distributed maps of some water quality parameters. A more complete indicator of erosion risk in a watershed is the multiple values of morphometric parameters, LULC, and sediment load. For planners and decision-makers to comprehend the morphological, LULC, and sediment load characteristics of any particular sub-watershed for planning at the sub-catchment level, AHP-VIKOR, GIS, and remote sensing approaches are more effective. LULC changes at a sub-watershed level by varying ranges of load had an impact on runoff and non-point source pollutant loading, including sediment, PO4 í, NO3 í , TP, and TN, as results revealed. The growth of built-up areas in response to the need for settlement and the rising change in agricultural land were the main causes of the increases in runoff volume, sediment, PO4 í, NO3 í , TP, and TN over two decades.
Description
Keywords
sub-watershed prioritization, water quality, surface water, Koka reservoir, upper Awash, morphometric parameters, VIKOR, remote sensing.