Search

Water management techniques to elevate rice yields and productive use of water resources in Africa, frequently lack a substantial spatial assessment as they are often based on plot level measurements without taking into account toposequential effects present in the landscape. These effects have been shown to significantly affect spatio-temporal variations in water availability and rice productivity in Asia. Therefore, this study addresses the spatio-temporal variations of the various water components within irrigated toposequences in an African inland valley and assesses its effect on water productivity and respective rice yields for two irrigation practices: (i) continuous flooding (CF), a well-known water management practice in rice cultivation used worldwide and (ii) a reduced irrigation scheme (RI) where irrigation is applied every 5 days resulting in a 1–2 cm water layer after irrigation. The lateral flow observed in the inland valley had a strong two-dimensional character, contributing to water gains between fields, located at the same toposequential level as well as along toposequences. The toposequential effect on sub-surface hydrological processes masked the overall effect of water management treatment on rice production. Additionally, the associated water productivity (WP) was not found to differ significantly between the treatments when standard calculations (i.e. net irrigation and evapotranspiration) were used but a clear toposequential effect was found for the fertilized lower lying fields when the net irrigation was corrected by the lateral flow component. Results of the established mixed regression model indicated that based on the groundwater table, rainfall and standard soil physico-chemical characteristics rice yields can be predicted in these African inland valleys under continuous flooding and reduced irrigation practices. Validation of the established regression function of inland valleys, representing various groundwater tables in the region, could lead to improved regression functions suitable to estimate spatial variation in rice production and water consumption across scales as affected by water management, fertilizer application and groundwater tables. [ABSTRACT FROM AUTHOR]

Highlights: [•] Inland valleys are common landscapes in Africa with an estimated area of 190Mha. [•] Inland valleys can significantly contribute to food security in sub-Saharan Africa. [•] 9.1% Of the total inland valley area could produce the current African rice demand. [•] The remaining 90.9% inland valley area could fulfil other ecosystem functions. [•] Seizing this regional inland valley potential requires a localized approach. [ABSTRACT FROM AUTHOR]

Abstract: Water productivity in agriculture needs to be improved significantly in the coming decades to secure food supply to a growing world population. To assess on a global scale where water productivity can be improved and what the causes are for not reaching its potential, the current levels must be understood. This paper describes the development and validation of a WATer PROductivity (WATPRO) model for wheat that is based on remote sensing-derived input data sets, and that can be applied at local to global scales. The model is a combination of Monteith''s theoretical framework for dry matter production in plants and an energy balance model to assess actual evapotranspiration. It is shown that by combining both approaches, the evaporative fraction and the atmospheric transmissivity, two parameters which are usually difficult to estimate spatially, can be omitted. Water productivity can then be assessed from four spatial variables: broadband surface albedo, the vegetation index NDVI, the extraterrestrial radiation and air temperature. A sensitivity analysis revealed that WATPRO is most sensitive to changes in NDVI and least sensitive to changes in air temperature. The WATPRO model was applied at 39 locations where water productivity was measured under experimental conditions. The correlation between measured and modelled water productivity was low, and this can be mainly attributed to differences in scales and in the experimental and modelling periods. A comparison with measurements from farmer''s fields in areas surrounded by other wheat fields located in Sirsa District, NW India, showed an improved correlation. Although not a validation, a comparison with SEBAL-derived water productivity in the same region in India proved that WATPRO can spatially predict water productivity with the same spatial variation. [Copyright &y& Elsevier]

Abstract: The growing pressure on fresh water resources demands that agriculture becomes more productive with its current water use. Increasing water productivity is an often cited solution, though the current levels of water productivity are not systematically mapped. A global map of water productivity helps to identify where water resources are productively used, and identifies places where improvements are possible. The WATPRO water productivity model for wheat, using remote sensing data products as input, was applied at a global scale with global data sets of the NDVI and surface albedo to benchmark water productivity of wheat for the beginning of this millennium. Time profiles of the NDVI were used to determine the time frame from crop establishment to harvest on a pixel basis, which was considered the modelling period. It was found that water productivity varies from approximately 0.2 to 1.8kg of harvestable wheat per cubic metre of water consumed. From the 10 largest producers of wheat, France and Germany score the highest country average water productivity of 1.42 and 1.35kgm−3, respectively. The results were compared with modelling information by who applied the GEPIC model at a global scale to map water productivity, and by who used FAO statistics to determine water productivity per country. A comparison with Liu et al. showed a good correlation for most countries, but the correlation with the results by Chapagain and Hoekstra was less obvious. The global patterns of the water productivity map were compared with global data sets of precipitation and reference evapotranspiration to determine the impact of climate and of water availability reflected by precipitation. It appears that the highest levels of water productivity are to be expected in temperate climates with high precipitation. Due to its non-linear relationship with precipitation, it is expected that large gains in water productivity can be made with in situ rain water harvesting or supplemental irrigation in dry areas with low seasonal precipitation. A full understanding of the spatial patterns by country or river basin will support decisions on where to invest and what measures to take to make agriculture more water productive. [Copyright &y& Elsevier]

Abstract: A methodology has been developed to quantify spatial variation of crop yield, evapotranspiration (ET) and water productivity (WPET) using the SEBAL algorithm and high and low resolution satellite images. SEBAL-based ET estimates were validated over an irrigated, wheat dominated area in the Yaqui Valley, Mexico and proved to be accurate (8.8% difference for 110 days). Estimated average wheat yields in Yaqui Valley of 5.5tha−1 were well within the range of measured yields reported in the literature. Measured wheat yields in 24 farmers’ fields in Sirsa district, India, were 0.4tha−1 higher than SEBAL estimated wheat yields. Area average WPET in the Yaqui Valley was 1.37kgm−3 and could be considered to be high as compared to other irrigated systems around the world where the same methodology was applied. A higher average WPET was found in Egypt''s Nile Delta (1.52kgm−3), Kings County (CA), USA (1.44kgm−3) and in Oldambt, The Netherlands (1.39kgm−3). The spatial variability of WPET within low productivity systems (CV=0.33) is higher than in high productivity systems (CV=0.05) because water supply in the former case is uncertain and farming conditions are sub-optimal. The high CV found in areas with low WPET indicates that there is considerable scope for improvement. The average scope for improvement in eight systems was 14%, indicating that 14% ET reduction can be achieved while maintaining the same yield. It is concluded that the proposed methodology is accurate and that better knowledge of the spatial variation of WPET provides valuable information for achieving local water conservation practices in irrigated wheat. [Copyright &y& Elsevier]

The great challenge of the agricultural sector is to produce more food from less water, which can be achieved by increasing Crop Water Productivity (CWP). Based on a review of 84 literature sources with results of experiments not older than 25 years, it was found that the ranges of CWP of wheat, rice, cotton and maize exceed in all cases those reported by FAO earlier. Globally measured average CWP values per unit water depletion are 1.09, 1.09, 0.65, 0.23 and 1.80 kg m-3 for wheat, rice, cottonseed, cottonlint and maize, respectively. The range of CWP is very large (wheat, 0.6–1.7 kg m-3; rice, 0.6–1.6 kg m-3; cottonseed, 0.41–0.95 kg m-3; cottonlint, 0.14–0.33 kg m-3 and maize, 1.1–2.7 kg m-3) and thus offers tremendous opportunities for maintaining or increasing agricultural production with 20–40% less water resources. The variability of CWP can be ascribed to: (i) climate; (ii) irrigation water management and (iii) soil (nutrient) management, among others. The vapour pressure deficit is inversely related to CWP. Vapour pressure deficit decreases with latitude, and thus favourable areas for water wise irrigated agriculture are located at the higher latitudes. The most outstanding conclusion is that CWP can be increased significantly if irrigation is reduced and crop water deficit is intendently induced. [Copyright &y& Elsevier]

Agricultural land suitability analysis (ALSA) for crop production is one of the key tools for ensuring sustainable agriculture and for attaining the current global food security goal in line with the Sustainability Development Goals (SDGs) of United Nations. Although some review studies addressed land suitability, few of them specifically focused on land suitability analysis for agriculture. Furthermore, previous reviews have not reflected on the impact of climate change on future land suitability and how this can be addressed or integrated into ALSA methods. In the context of global environmental changes and sustainable agriculture debate, we showed from the current review that ALSA is a worldwide land use planning approach. We reported from the reviewed articles 69 frequently used factors in ALSA. These factors were further categorized in climatic conditions (16), nutrients and favorable soils (34 of soil and landscape), water availability in the root zone (8 for hydrology and irrigation) and socio-economic and technical requirements (11). Also, in getting a complete view of crop's ecosystems and factors that can explain and improve yield, inherent local socio-economic factors should be considered. We showed that this aspect has been often omitted in most of the ALSA modeling with only 38% of the total reviewed article using socio-economic factors. Also, only 30% of the studies included uncertainty and sensitivity analysis in their modeling process. We found limited inclusions of climate change in the application of the ALSA. We emphasize that incorporating current and future climate change projections in ALSA is the way forward for sustainable or optimum agriculture and food security. To this end, qualitative and quantitative approaches must be integrated into a unique ALSA system (Hybrid Land Evaluation System - HLES) to improve the land evaluation approach. • Methods used in agricultural land suitability analysis (ALSA) strengths and limitations • Uncertainty and sensitivity analysis inclusion in ALSA modeling process • Climate change analysis inclusion in ALSA task • Frequently used predictors in ALSA modeling including biophysical, socioeconomic and management practices [ABSTRACT FROM AUTHOR]

Mangrove swamp rice cultivation is important for food security in some countries of West Africa including Sierra Leone. In this agro-ecology, rice is cultivated during the rainy season when freshwater flows in the rivers and salt and acidity concentrations have reduced to non-toxic levels. Rice yields in the mangrove ecosystem of Sierra Leone are higher than in other agro-ecologies and weed, disease and pest pressures are minimal. However, salinity, acidity and crabs negatively affect rice productivity in the mangrove swamps. Due to the differences in levels of flooding, salinity and acid sulphate conditions of mangrove swamp soils, it is assumed that there is variability of soil properties of mangrove swamps along the associated river, which may impact the choice of suitable rice varieties and soil management practices. The purpose of this study was to understand the soil physical and chemical properties of mangrove swamp soils along the Great Scarcies River of Sierra Leone. A soil sampling survey was designed and implemented using transects to collect composite soil samples of 1 ha area at 0–0.2 m depth at 11 different sites located from the estuary of the Great Scarcies River to approximately 35 km inland. The soil samples were air-dried, processed and analyzed for selected physical and chemical properties by recommended methods. Statistical analysis generated mean, standard deviations, coefficient of variation, correlation matrix and principal components. The high variability in soil physical and chemical characteristics of mangrove swamp soils along the Great Scarcies River could be attributed to the complex interactions between the twice daily tidal inundations and depositions of soil organic matter, physical particles and nutrients onto the mangrove swamp soils along the river. The result of this is a soil fertility gradient down-stream. [ABSTRACT FROM AUTHOR]

Achieving sustainable agricultural intensification in inland valleys while limiting the impacts on water quantity and water quality requires a better understanding of the valleys’ hydrological behavior with respect to their contributing watersheds. This study aims at assessing the dynamics of hydrological processes and nitrate loads within inland valleys that are experiencing different land uses. To achieve this goal, an HRU-based interface (ArcSWAT2012) and a grid-based setup (SWATgrid) of the Soil Water Assessment Tool (SWAT) model were applied to three headwater inland valley watersheds located in the commune of Djougou in central Benin that are characterized by different proportions of cultivated area. Satisfactory model performance was obtained from the calibration and validation of daily discharges with the values of R 2 and NSE mostly higher than 0.5, but not for nitrate loads. The annual water balance reveals that more than 60% of precipitation water is lost to evapotranspiration at all sites, amounting to 868 mm in Kounga, 741 mm in Tossahou, and 645 mm in Kpandouga. Percolation (302 mm) is important in the Kpandouga watershed which is dominated by natural vegetation at 99.7%, whereas surface runoff (105 mm) and lateral flow (92 mm) are the highest in the Kounga watershed having the highest proportion of agricultural land use (14%). In all the studied watersheds, nitrate loads are very low (not exceeding 4000 KgN per year) due to the low fertilizer application rates, and the water quality is not threatened if a standard threshold of 10 mg/l NO 3 -N is applied. The results achieved in this study show that SWAT can successfully be used in spatial planning for sustainable agricultural development with limited environmental impact on water resources in inland valley landscapes. [ABSTRACT FROM AUTHOR]

Achieving rice self-sufficiency in West Africa will require an expansion of the irrigated rice area under water-scarce conditions. However, little is known about how much area can be irrigated and where and when water-saving practices could be used. The objective of this study was to assess potentially irrigable lands for irrigated rice cultivation under water-saving technology in Burkina Faso. A two-step, spatially explicit approach was developed and implemented. Firstly, machine learning models, namely Random Forest (RF) and Maximum Entropy (MaxEnt) were deployed in ecological niche modeling (ENM) approach to assess the land suitability for irrigated rice cultivation. Spatial datasets on topography, soil characteristics, climate parameters, land use, and water were used along with the current distribution of irrigated rice locations in Burkina Faso to drive ENMs. Secondly, the climatic suitability for alternate wetting and drying (AWD), an irrigation management method for saving water in rice cultivation in irrigated systems, was assessed by using a simple water balance model for the two main growing seasons (February to June and July to November) on a dekadal time scale. The evaluation metrics of the ENMs such as the area under the curve and percentage correctly classified showed values higher than 80% for both RF and MaxEnt. The top four predictors of land suitability for irrigated rice cultivation were exchangeable sodium percentage, exchangeable potassium, depth to the groundwater table, and distance to stream networks and rivers. Potentially suitable lands for rice cultivation in Burkina Faso were estimated at 21.1 × 105 ha. The whole dry season was found suitable for AWD implementation against 25–100% of the wet season. Soil percolation was the main driver of the variation in irrigated land suitability for AWD in the wet season. The integrated modeling and water balance assessment approach used in this study can be applied to other West African countries to guide investment in irrigated rice area expansion while adapting to climate change. • The potential for irrigated rice in Burkina Faso was estimated at 21.10 × 105 ha. • All dekads in the dry season are suitable for Alternate Wetting and Drying (AWD). • About 25–100% of wet season dekads are suitable for AWD. • Soil percolation was the main driver of land suitability to AWD in the wet season. [ABSTRACT FROM AUTHOR]

• Evaluation of 12 gridded actual evaporation datasets for hydrological modelling. • Comparison of four distinct model calibration strategies. • Process-diagnostic of model outputs with multiple independent variables. • Calibration only on spatial patterns of evaporation improves the model responses. • Model calibration strategy determines the utility of the evaporation data. Twelve actual evaporation datasets are evaluated for their ability to improve the performance of the fully distributed mesoscale Hydrologic Model (mHM). The datasets consist of satellite-based diagnostic models (MOD16A2, SSEBop, ALEXI, CMRSET, SEBS), satellite-based prognostic models (GLEAM v3.2a, GLEAM v3.3a, GLEAM v3.2b, GLEAM v3.3b), and reanalysis (ERA5, MERRA-2, JRA-55). Four distinct multivariate calibration strategies (basin-average, pixel-wise, spatial bias-accounting and spatial bias-insensitive) using actual evaporation and streamflow are implemented, resulting in 48 scenarios whose results are compared with a benchmark model calibrated solely with streamflow data. A process-diagnostic approach is adopted to evaluate the model responses with in-situ data of streamflow and independent remotely sensed data of soil moisture from ESA-CCI and terrestrial water storage from GRACE. The method is implemented in the Volta River basin, which is a data scarce region in West Africa, for the period from 2003 to 2012. Results show that the evaporation datasets have a good potential for improving model calibration, but this is dependent on the calibration strategy. All the multivariate calibration strategies outperform the streamflow-only calibration. The highest improvement in the overall model performance is obtained with the spatial bias-accounting strategy (+29%), followed by the spatial bias-insensitive strategy (+26%) and the pixel-wise strategy (+24%), while the basin-average strategy (+20%) gives the lowest improvement. On average, using evaporation data in addition to streamflow for model calibration decreases the model performance for streamflow (-7%), which is counterbalance by the increase in the performance of the terrestrial water storage (+11%), temporal dynamics of soil moisture (+6%) and spatial patterns of soil moisture (+89%). In general, the top three best performing evaporation datasets are MERRA-2, GLEAM v3.3a and SSEBop, while the bottom three datasets are MOD16A2, SEBS and ERA5. However, performances of the evaporation products diverge according to model responses and across climatic zones. These findings open up avenues for improving process representation of hydrological models and advancing the spatiotemporal prediction of floods and droughts under climate and land use changes. [ABSTRACT FROM AUTHOR]

Water availability and high soil fertility make inland valley landscapes suitable for sustainable rice-based cropping. In this study, Random Forests statistical analysis was used on a database of 499 surveyed inland valleys in four study zones in three West African countries. The goal of the study was to assess parameters that indicate (are predictors for) high potential for development of rice-based systems in inland valleys. These parameters are related to the biophysical (hydrology, soil, climate, and topography) and socio-economic (demography, accessibility, and markets) environments. Farmer group surveys and secondary data from existing publicly available spatial data sets were used. The analysis revealed that, across the four research areas, the following parameters were relevant predictors for rice development: (1) distance from the inland valley to the nearest market; (2) distance from the inland valley to the nearest rice mill; (3) population density in the immediate environment of the inland valley; (4) total nitrogen in the top 20 cm of the soil profile; (5) land elevation; and (6) soil texture on the upper slope of the inland valley. Several predictors were highly important for specific research areas, but not for all, thus showing the diversity in the studied agricultural landscapes. These predictors included soil fertility management, source of irrigation water, and the percentage of female farmers in the inland valley. The identified relevant predictors will be used to map the potential rice production development of the inland valleys. This will help development agencies to assess their zones based on quantitative analysis for inland valley potential development. [ABSTRACT FROM AUTHOR]

Identifying the biophysical factors that affect the performance of irrigated crops in semi-arid conditions is pivotal to the success of profitable and sustainable agriculture under variable climate conditions. In this study, soil physical and chemical variables and plots characteristics were used through linear mixed and random forest-based modeling to evaluate the determinants of actual evapotranspiration (ET a) and crop water productivity (CWP) in rice in the Kou Valley irrigated scheme in Burkina Faso. Multi-temporal Landsat images were used within the Python module for the Surface Energy Balance Algorithm for Land model to calculate rice ET a and CWP during the dry seasons of 2013 and 2014. Results showed noticeable spatial variations in PySEBAL-derived ET a and CWP in farmers' fields during the study period. The distance between plot and irrigation scheme inlet (D PSI), plot elevation, sand and silt contents, soil total nitrogen, soil extractable potassium and zinc were the main factors affecting variabilities in ET a and CWP in the farmers' fields, with D PSI being the top explanatory variable. There was generally a positive association, up to a given threshold, between ET a and D PSI , sand and silt contents and soil extractable zinc. For CWP the association patterns for the top six predictors were all non-monotonic; that is a mix of increasing and decreasing associations of a given predictor to either an increase or a decrease in CWP. Our results indicate that improving irrigated rice performance in the Kou Valley irrigation scheme would require growing more rice at lower altitudes (e.g. < 300 m above sea level) and closer to the scheme inlet, in conjunction with a good management of nutrients such as nitrogen and potassium through fertilization. • Biophysical factors affecting irrigation rice performance were assessed. • Landsat images were used to estimate ET a and CWP. • Large spatial variations were observed in ET a and CWP in farmers' fields. • Soil properties and plots position and elevation explained variabilities in ET a and CWP. • Improved plot and nutrient management can increase irrigated rice performance. [ABSTRACT FROM AUTHOR]

The selection of optimal areas for specific cultivation systems is an important step in achieving increased, sustainable rice production in Benin. This study aims to determine suitable areas for rice production in the inland valley of Tossahou using a GIS-based approach that evaluates and combines biophysical factors such as climate, hydrology, soil and landscape, following the FAO parameter method and guidelines for land evaluation. Soil and landscape suitability was assessed for three different rice cultivation systems: rainfed bunded (RB), cultivation under natural flooding (NF), and irrigated cultivation (RI). The results show that in the inland valley (mostly including the hydromorphic zones and the valley bottom) 52% of the area is suitable for irrigated cultivation, 18% for cultivation under natural flood and 1.2% for rainfed bunded rice. Precipitation and temperature were limiting factors for all cultivation systems. Flooding was the most limiting factor for NF while RI and RB were mostly limited by steep slopes and soil texture respectively. As a first attempt in Benin, this study can play an important role in achieving optimised rice production in inland valleys, and additional studies including socio-economic aspects, carried out in the same area, or in areas under similar conditions, are relevant to close the yield gap and improve the selection approach. [ABSTRACT FROM AUTHOR]

Rice is one of the major staple foods in sub-Saharan Africa (SSA) and is mainly grown in three environments: rainfed upland and rainfed and irrigated lowlands. In all rice-growing environments, the yield gap (the difference between the potential yield in irrigated lowland or water-limited yield in rainfed lowland and upland and the actual yield obtained by farmers) is largely due to a wide range of constraints including water-related issues. This paper aims to review water management research for rice cultivation in SSA. Major water-related constraints to rice production include drought, flooding, iron toxicity, and soil salinity. A wide range of technologies has been tested by Africa Rice Center (AfricaRice) and its partners for their potential to address some of the water-related challenges across SSA. In the irrigated lowlands, the system of rice intensification and alternate wetting and drying significantly reduced water use, while the pre-conditions to maintain grain yield and quality compared to continuous flooding were identified. Salinity problems caused by the standing water layer could be addressed by flushing and leaching. In the rainfed lowlands, water control structures, Sawah rice production system, and the Smart-Valleys approach for land and water development improved water availability and grain yield compared to traditional water management practices. In the rainfed uplands, supplemental irrigation, mulching, and conservation agriculture mitigated the effects of drought on rice yield. The Participatory Learning and Action Research (PLAR) approach was developed to work with and educate communities to help them implement improved water management technologies. Most of the research assessed a few indicators such as rice yield, water use, water productivity at the field level. There has been limited research on the cost-benefit of water management technologies, enabling conditions and business models for their large-scale adoption, as well as their impact on farmers' livelihoods, particularly on women and youth. Besides, limited research has been conducted on water management design for crop diversification, landscape-level water management, and iron toxicity mitigation, particularly in lowlands. Filling these research gaps could contribute to sustainable water resources management and sustainable intensification of rice-based systems in SSA. • Research on water management for rice cultivation in sub-Saharan Africa was reviewed. • The majority of studies aimed at improving water control in rainfed lowlands. • The studies focused mostly on improving rice yield and water productivity. • Less attention was given to cost-benefits and impact of water control technologies. • Water management was mostly studied at the field- rather than at landscape-level. [ABSTRACT FROM AUTHOR]

Although rice production has increased significantly in the last decade in West Africa, the region is far from being rice self-sufficient. Inland valleys (IVs) with their relatively higher water content and soil fertility compared to the surrounding uplands are the main rice-growing agroecosystem. They are being promoted by governments and development agencies as future food baskets of the region. However, West Africa's crop production is estimated to be negatively affected by climate change due to the strong dependence of its agriculture on rainfall. The main objective of the study is to apply a set of machine learning models to quantify the extent of climate change impact on land suitability for rice using the presence of rice-only data in IVs along with bioclimatic indicators. We used a spatially explicit modeling approach based on correlative Ecological Niche Modeling. We deployed 4 algorithms (Boosted Regression Trees, Generalized Linear Model, Maximum Entropy, and Random Forest) for 4-time periods (the 2030s, 2050s, 2070s, and 2080s) of the 4 Representative Concentration Pathways (RCP2.6, RCP4.5, RCP6.0, and RCP8) from an ensemble set of 32 spatially downscaled and bias-corrected Global Circulation Models climate data. The overall trend showed a decrease in suitable areas compared to the baseline as a function of changes in temperature and precipitation by the order of 22–33% area loss under the lowest reduction scenarios and more than 50% in extreme cases. Isothermality or how large the day to night temperatures oscillate relative to the annual oscillations has a large impact on area losses while precipitation increase accounts for most of the areas with no change in suitability. Strong adaptation measures along with technological advancement and adoption will be needed to cope with the adverse effects of climate change on inland valley rice areas in the sub-region. The demand for rice in West Africa is huge. For the rice self-sufficiency agenda of the region, "where" and "how much" land resources are available is key and requires long-term, informed planning. Farmers can only adapt when they switch to improved breeds, providing that they are suited for the new conditions. Our results stress the need for land use planning that considers potential climate change impacts to define the best areas and growing systems to produce rice under multiple future climate change uncertainties. [Display omitted] • West Africa's agenda for rice self-sufficiency remains uncertain under climate change conditions. • The potential of inland valleys for rainfed rice cultivation under future climate change uncertainties was assessed. • Significant losses of suitable areas due to changes in the day to night temperatures oscillations were found. [ABSTRACT FROM AUTHOR]

Small reservoirs are one of the most important sources of water for irrigation, domestic and livestock uses in the Upper East Region (UER) of Ghana. Despite various studies on small reservoirs in the region, information on their spatial-temporal variations is minimal. Therefore, this study performed a binary Random Forest classification on Sentinel-2 images for five consecutive dry seasons between 2015 and 2020. The small reservoirs were then categorized according to landscape positions (upstream, midstream, and downstream) using a flow accumulation process. The classification produced an average overall accuracy of 98% and a root mean square error of 0.087 ha. It also indicated that there are currently 384 small reservoirs in the UER (of surface area between 0.09 and 37 ha), with 20% of them newly constructed between the 2016-17 and 2019-20 seasons. The study revealed that upstream reservoirs have smaller sizes and are likely to dry out during the dry season while downstream reservoirs have larger sizes and retain substantial amounts of water even at the end of the dry season. The results further indicated that about 78% of small reservoirs will maintain an average of 54% of their water surface area by the end of the dry season. This indicates significant water availability which can be effectively utilized to expand dry season irrigation. Overall, we demonstrate that landscape positions have significant impact on the spatial-temporal variations of small reservoirs in the UER. The study also showed the effectiveness of remote sensing and machine learning algorithms as tools for monitoring small reservoirs. • There are currently 384 small reservoirs in the Upper East Region of Ghana. • Twenty percent of small reservoirs were constructed between the 2016-17 and 2019-20 seasons. • Upstream reservoirs have smaller sizes and are likely to dry out in the dry season. • Downstream reservoirs have larger sizes and retain substantial amounts of water by the end of the dry season. • About 78% of small reservoirs will maintain an average of 54% of their water surface area by the end of the dry season. [ABSTRACT FROM AUTHOR]