Search

Climate variability poses great challenges to food security in West Africa, a region heavily dependent on rainfall for farming. Identifying sowing strategies that minimize yield losses for farmers in the region is crucial to securing their livelihood. In this paper, we investigate three sowing strategies to assess their ability to identify safe sowing windows for smallholder farmers in the Sudanian region of West Africa (WA) in the context of a changing climate. The GIS version of the FAO crop model, AquaCrop-GIS, is used to simulate the yield response of maize (Zea mays L.) to varying sowing dates throughout the rainy season across WA. Based on an average of 38 years of data per grid cell, we identify safe sowing windows across the Sudanian region that secure at least 90% of maximal yield. We find that current sowing strategies, based on minimum thresholds for rainfall accumulated over a period that are widely applied in the region, carry a higher risk of yield failure, especially at the beginning of the rainy season. This analysis shows that delaying sowing for a month to mid-June in the central region (east of Lon 8.5°W), and to early August in the semi-arid areas is a safer strategy that ensures optimal yields. A comparison between the periods 1982–1991 and 1992–2019 shows a negative shift for LO10 mm and LO20 mm, suggesting a wetter regime compared to the dry periods of the 1970s and 1980s. On the contrary, we observe a positive shift in the safe window strategy, highlighting the need for precautions due to erratic rainfall at the beginning of the season. The precipitation-based strategies hold a high risk, while the safe sowing window strategy, easily accessible to smallholder farmers, is more fitting, given the current climate. [ABSTRACT FROM AUTHOR]

West African food systems and rural socio-economics are based on rainfed agriculture, which makes society highly vulnerable to rainfall uncertainty and frequent floods and droughts. Reliable rainfall information is currently missing. There is a sparse and uneven rain gauge distribution and, despite continuous efforts, rainfall satellite products continue to show weak correlations with ground measurements. This paper aims to investigate whether water vapor (WV) observations together with temporal information can complement thermal infrared (TIR) data for satellite rainfall retrieval in a Deep Learning (DL) framework. This is motivated by the fact that water vapor plays a key role in the highly seasonal West African rainfall dynamics. We present a DL model for satellite rainfall detection based on WV and TIR channels of Meteosat Second Generation and temporal information. Results show that the WV inhibition of low-level features enables the depiction of strong convective motions usually related to heavy rainfall. This is especially relevant in areas where convective rainfall is dominant, such as the tropics. Additionally, WV data allow us to detect dry air masses over our study area, that are advected from the Sahara Desert and create discontinuities in precipitation events. The developed DL model shows strong performance in rainfall binary classification, with less false alarms and lower rainfall overdetection (FBias < 2.0 ) than the state-of-the-art Integrated MultisatellitE Retrievals for GPM (IMERG) Final Run. [ABSTRACT FROM AUTHOR]

Food and economic security in West Africa rely heavily on rainfed agriculture and are threatened by climate change and demographic growth. Accurate rainfall information is therefore crucial to tackling these challenges. Particularly, information about the occurrence and length of droughts as well as the onset date of the rainy season is essential for agricultural planning. However, existing rainfall models fail to accurately represent the highly variable and sparsely monitored West African rainfall patterns. In this paper, we show the potential of deep learning (DL) to model rainfall in the region and propose a methodology to develop DL models in data-scarce areas. We built two DL models for satellite rainfall (rain/no-rain) detection over northern Ghana from Meteosat TIR data based on standard DL architectures: Convolutional neural networks (CNNs) and convolutional long short-term memory neural networks (ConvLSTM). The Integrated Multi-satellitE Retrievals for the Global Precipitation Measurement (GPM) mission (IMERG) and Precipitation Estimation from Remotely Sensed Imagery Using an Artificial Neural Network Cloud Classification System (PERSIANN-CCS) products are used as benchmarks. We use rain gauge data from the Trans-African Hydro-Meteorological Observatory (TAHMO) for model development and performance evaluation. We show that our models compare well against existing products despite being considerably simpler, developed with a small training dataset—i.e., 8 stations covering 2.5 years with 20.4% of the data missing—and using TIR data alone. Concretely, our models consistently outperform PERSIANN-CCS for rain/no-rain detection at a sub-daily timescale. While IMERG is the overall best performer, the DL models perform better in the second half of the rainy season despite their simplicity (i.e., up to 120 k parameters). Our results suggest that DL-based regional models are a promising alternative to state-of-the-art global products for providing regional rainfall information, especially in meteorologically complex regions such as the (sub)tropics, which are poorly covered by ground-based rainfall observations. [ABSTRACT FROM AUTHOR]

Abstract: This article provides an overview of both experimental and modeling research carried out over the past 15 years by the authors addressing scaling effects in Hortonian surface runoff. Hortonian surface runoff occurs when rainfall intensity exceeds infiltration capacity of the soil. At three sites in West Africa (Côte d’Ivoire, Ghana, and Burkina Faso) runoff was measured from plots of different lengths to assess scale effects. Consistently, longer plots showed much lower runoff percentages than shorter plots. There were large variations in runoff percentages from one rainstorm to the next but there were very good correlations between plots of equal length for each single event. This strongly suggests that temporal dynamics are the cause behind the observed scale effects. In the literature, spatial variability is often proffered as explanation for such scale effects without providing a mechanism that would cause consistent reduction in runoff percentages with increasing slope length. To further examine whether temporal dynamics can indeed provide the explanation, Hortonian runoff was simulated using models with increasing levels of complexity. The simplest model was already able to reproduce the observed scale effects. Also more complex models were used that accounted explicitly for spatial variability. The conclusions remained the same regarding the role of temporal dynamics. Finally, a dimensional analysis was developed that helps predict under which circumstances one can expect scale effects similar to the ones observed in West Africa. [Copyright &y& Elsevier]

Impacts of climate change vary from region to region. The 4th Assessment Report of the Intergovernmental Panel on Climate Change (IPCC) mentions that drier areas will be affected by more droughts while the rainfall regime, in general, will become 'rougher'. In West Africa, specifically the area below the Sahel, the climate change signal may be more subtle. Anecdotal evidence from farmers suggests that the onset of rainy season has been shifting forward in time over the past two generations. Recently, detailed atmospheric modelling over the region shows that in the near future too, the onset of rainy season will shift to later periods in the year, roughly from April towards May. The end of rainy season as well as the total amount of rainfall will remain more or less fixed. This implies that adaptation strategies should be twofold. The first part of a comprehensive adaptation strategy would be a continuation of the efforts to produce faster growing rainfed crop cultivars, mainly corn and sorghum. The second part would consist of increased water storage during the wet season for use during dry season. [ABSTRACT FROM AUTHOR]

In semiarid regions, thousands of small reservoirs provide the rural population with water, but their storage volumes and hydrological impact are largely unknown. This paper analyzes the suitability of weather-independent radar satellite images for monitoring small reservoir surfaces. The surface areas of three reservoirs were extracted from 21 of 22 ENVISAT Advanced Synthetic Aperture Radar scenes, acquired bimonthly from June 2005 to August 2006. The reservoir surface areas were determined with a quasi-manual classification approach, as stringent classification rules often failed due to the spatial and temporal variability of the backscatter from the water. The land—water contrast is critical for the detection of water bodies. Additionally, wind has a significant impact on the classification results and affects the water surface and the backscattered radar signal (Bragg scattering) above a wind speed threshold of 2.6 m · s-1. The analysis of 15 months of wind speed data shows that, on 96% of the days, wind speeds were below the Bragg scattering criterion at the time of night time acquisitions, as opposed to 50% during the morning acquisition time. Night time acquisitions are strongly advisable over day time acquisitions due to lower wind interference. Over the year, radar images are most affected by wind during the onset of the rainy season (May and June). We conclude that radar and optical systems are complimentary. Radar is suitable during the rainy season but is affected by wind and lack of vegetation context during the dry season. [ABSTRACT FROM AUTHOR]

In West Africa, which is an extremely moisture-limited region, soil water information plays a vital role in hydrologic and meteorologic modeling for improved water resource planning and food security. Recent and upcoming satellite missions, such as SMOS and MetOp, hold promise for the regional observation of soil moisture. The resolution of the satellites is relatively coarse (> 100 km²), which brings with it the need for large-scale soil moisture information for calibration and validation purposes. We put forward a soil moisture sampling protocol based on hydrotopes. Hydrotopes are defined as landscape units that show internally consistent hydrologic behavior. This hydrotope analysis helps in the following ways: 1) by ensuring statistically reliable validation via the reduction of the overall pixel variance and 2) by improving sampling schemes for ground truthing by reducing the chance of sampling bias. As a sample application, we present data from three locations with different moisture regimes `within the Volta Basin during both dry and wet periods. Results show that different levels of reduction in the overall pixel variance of soil moisture are obtained, depending on the general moisture status. With respect to the distinction between the different hydrotope units, it is shown that under intermediate moisture `conditions, the distinction between the different hydrotope units is highest, whereas extremely dry or wet conditions tend to have a homogenizing effect on the spatial soil moisture distribution. This paper confirms that well-defined hydrotope units yield an improvement at pixel-scale soil moisture averages that can easily be applied. [ABSTRACT FROM AUTHOR]

Abstract: Measurements of surface runoff from uniform slopes of different lengths in West Africa have shown that longer slopes tend to have less runoff per unit of length than short slopes. The main reason for this scale effect is that once the rain stops, water on long slopes has more opportunity time to infiltrate than water on short slopes. A validated simple model is put forward that quantifies the scale effects and predicts under which circumstances they can be expected to be most significant. An overview is given of management options that are available to farmers to reduce runoff at the field and slope level. The potential effects of these management options on the hydraulic conductivity, Ksat, and Manning''s roughness coefficient, n, are qualified. The model is subsequently used to calculate the effect of changes in Ksat and n on the runoff from slopes under rainfall conditions found in West Africa. Finally, two design examples are given, namely maximum field lengths and dimensioning of interceptor drains. [Copyright &y& Elsevier]

Summary: Seasonal variability of the energy partitioning was analyzed with a combination of eddy fluxes of sensible and latent heat and weather data on intensely farmed land in the savanna area of Eastern Burkina Faso, West Africa. The analysis covers two rainy seasons (May–October 2003 and 2004), one dry season (December–April 2004), one dry to wet transition period (May–June 2004) and two wet to dry transition periods (October–November 2003 and 2004). This is the first long-term flux measurement reported upon for this part of the world. The inter-annual partitioning of the available energy between the latent heat (actual evaporation) and sensible heat flux shows a conservative relationship between a decrease in annual rainfall (33%) and the sum of the increase in sensible heat flux (20%) and decrease in latent heat flux (10%). The latent heat flux was the main consumer of the available energy during the rainy season (71%), while sensible heat flux was dominant during the dry season (77%). In terms of the rainy season water balance, 1229mm of rainfall was observed in 2003, of which 351mm evaporated, while 825mm of rainfall was observed in 2004, of which 268mm evaporated. During the dry season, the latent heat flux was strongly coupled to the atmosphere, with the decoupling coefficient (Ω) ranging from 0.18 to 0.4. These low values were due to low soil moisture availability and high vapor pressure deficit (VPD>4kPa). During the rainy season, the latent heat flux was decoupled from the atmosphere (0.6⩽ Ω <0.9). Following this pattern of coupling and decoupling, a new formulation of actual evaporation and surface conductance was proposed. The results are thought to be relevant inputs for eco-hydrological models in the semi-arid region of West Africa. [Copyright &y& Elsevier]