Regional to continental modelling of water quality in Africa: Imprints of global change on river nutrients and sediments

Water pollution is a global challenge threatening human health, limiting food production and
reducing ecosystem functions. The situation has worsened in almost all rivers in the Global
South since the 1990s. However, Africa’s aquatic ecosystems have not been the focal point of
research in comparison to other global regions. In many parts of Africa, land use changes such
as agricultural expansion and intensification, rapid urbanization and industrialization are
degrading the water quality of lakes and rivers, thereby threatening the sustainable use of water
resources. At the same time, climate change adds a new uncertainty with regards to the water
quality of resources on the continent. In this study, we advance our understanding of water
quality, with regards to river nutrients and sediment export for Africa under several global
change drivers. To achieve this, we utilize the Soil and Water Assessment Tool (SWAT+)
water quality model, setup for regional to continental scale application.
With agriculture identified as the major contributor to nonpoint source (NPS) pollution of water
resources, this research initially focuses on improving the representation of agricultural land
use and crop management within the SWAT+ model for tropical and subtropical regions. Leaf
Area Index (LAI) and evapotranspiration (ET) obtained from Earth Observations are used as
proxies to evaluate crop phenology and cropping seasons. The results for the Nile basin show
improved estimates of LAI and ET. These improvements subsequently translate into more
accurate assessments of soil erosion, given the key influence of crop cover on cropland
erodibility.
Next, this research proposes a general protocol that should be incorporated in the SWAT+ large
scale soil erosion modelling framework, especially in data-scarce regions utilizing existing
global datasets. The proposed protocol includes: (a) incorporating topographic factors from
high/medium resolution Digital Elevation Models (DEMs), (b) incorporating crop phenology
data, (c) introducing an areal threshold to linearize sediment yield in large model units and (d)
applying a hydrological mass balance calibration. This workflow is tested in a data-scare region
(Nile basin) under climate change projections utilizing existing global datasets. Results show
that an improved regionalisation protocol of soil erosion influences the projected soil erosion
and river sediment loads under climate change.
Subsequently, this study focuses on the historical climate impact attribution of river water
quantity and quality changes. We adopt the Inter-Sectoral Impact Model Intercomparison
Project (ISIMIP3a) climate impact attribution approach and apply it in the Nile basin focusing
on long-term temporal changes in river flows and river sediment loads. Results show that the
headwaters of the White Nile, particularly the Lake Victoria basin, have experienced a trend
towards wetter conditions, leading to increased river flows and heightened risks of soil erosion,
further intensifying river sediment loads. In contrast, the impact of climate change on river
flows and sediment loads is comparatively smaller in the Blue Nile basin and the Main Nile
basin. This is primarily due to the smaller magnitudes of changes in mean annual precipitation
in these regions. These findings show the spatial differences in the impacts of climate on river
flows and sediment loads while highlighting the most impacted region in the basin.
Later on, we setup a continental water quality model to map out Total Nitrogen (TN) and Total
Phosphorus (TP) hotspots for the recent decade (2010 – 2019). Utilizing the Sustainable
Development Goal (SDG) 6.3.2 criteria, which designates a water body as having "good
ambient water quality" if 80 % or more of its monitored values meet their targets, our findings
show that 45 % and 14 % of African rivers fail to meet the set water quality thresholds for TP
and TN, respectively. When synthesizing data for both TP and TN, 33 % of the rivers do not
qualify as having “good ambient water quality”. Geographically, the most pronounced nutrient
hotspots are identified in North Africa, the Niger Delta in West Africa, specific zones in
Southern Africa, and the Nile basin. Broadly, these regions overlap with areas of high fertilizer
and manure inputs, and high point source loads. Therefore, implementing targeted strategies
for efficient fertilizer application and wastewater treatment is crucial in Africa to address the
high nutrient concentrations especially for TP.
Finally, we look at the historical impact of human and climate drivers on TN and TP exports
to African coastal waters over the past 40 years (1980 to 2019). Our analysis focuses on four
major drivers: (1) climate change, (2) fertilizer and manure changes, (3) point source changes,
and (4) land use changes. Results show that climate change has been the dominant driver,
causing both increases and decreases in nutrient exports but overall leading to a 15 % reduction
in TN and an 8 % decrease in TP exports, primarily due to reduced runoff resulting from
decreased precipitation. Contrary, fertilizer and manure application, along with point source
loadings, have increased TN and TP export by 6 % and 1 % respectively. Land use changes
have had negligible effects (< ± 0.1 %). These insights show the differing impacts of global
change drivers on nutrient export. While climate change is having mixed impacts, human
drivers are primarily contributing to an increase in the nutrient fluxes.
Overall, this study serves the dual purpose of addressing the research gap in regional to
continental modelling of river nutrient water pollution and attributing the impact of global
change drivers on river nutrient water quality. By doing so, we aim to bridge the substantial
water quality data gap in Africa, particularly for river sediments, TN and TP. This water quality
model can serve as a valuable tool for informing assessments such as SDGs, across a wide
spectrum of water quality issues, spanning from inland to coastal systems. However, the
significance of this work extends beyond the realm of water quality modelling. It holds
relevance for the broader scientific community, offering essential methodological approaches
and valuable data that can influence policies related to water resources management.