1. Introduction
Precipitation in the semiarid Sahel region has a large interannual-to-decadal variability and a high vulnerability to hydrologic extremes. Despite the exponentially growing number of scientific studies of Sahelian precipitation (Rodríguez-Fonseca et al. 2015), there is still a large uncertainty about the most prominent questions: 1) what are the sources of moisture causing fluctuations in Sahelian precipitation, 2) what is the role of changes in vegetation resulting from deforestation, desertification, or recently observed regreening, and 3) what is the influence of global climate change on the availability of water?
For this last question, Monerie et al. (2021) state that the magnitude of the future change in precipitation in the Sahel is uncertain and there is only an agreement on its sign. That is, there will be an increase in summer precipitation over the eastern Sahel and a decrease over the western Sahel. There is an agreement on the positive vegetation–rainfall feedback first proposed by Charney (1975). However, experiments investigating the albedo, moisture, and momentum feedbacks draw different conclusions (Yu et al. 2017). The Gulf of Guinea and the eastern part of the Atlantic Ocean were recognized as the major source of moisture for the region, associated with the West African monsoon (Thorncroft et al. 2011). Nieto et al. (2006) identified sources of moisture also in the North Atlantic Ocean between the Sahel and Iberia, the Mediterranean basin, and the Red Sea. On the other hand, Yu et al. (2017) conclude that the positive vegetation–rainfall feedbacks are dominated by moisture recycling. They emphasize the limitations in the simulations driven by the sea surface temperature (SST) anomalies for the prediction of the apparent recovery from the multidecadal drought after the year 2000. In summary, vegetation and vegetation changes appear to be the key points dominating the above questions.
Large changes in the Sahelian vegetation are planned with the pan-African Great Green Wall for the Sahara and the Sahel initiative (GGW). The GGW is a reforestation program to reverse the degradation of the land from Senegal to Djibouti, involving a wide range of stakeholders (Goffner et al. 2019).
Saley et al. (2019) investigated the potential impact of a hypothetical GGW, employing the limited-area climate model RegCM. In a scenario model run, they changed the land cover within a band between 14.08° and 15.84°N latitude from short grass to deciduous needleleaf trees. The results indicated that such a green belt would increase the number of rainy days in the Sahel by +9%. From a reforestation study, Oguntunde et al. (2014) argued that reforestation reduces the projected warming and drying over the reforested zones by reducing the speed of the monsoon flow. On the other hand, in the summer, this delays the onset of the monsoon and transport of cool moist air to regions located downwind from the reforested region. In consequence, these areas suffer from temperature increases as well as from reduced rainfall. Hypothetical reforestation modeling studies performed by Bamba et al. (2019) showed that the location of the potential area precipitation increases depended on the latitude of the forested areas. Rodríguez-Fonseca et al. (2015) derived consistent results across models on the sensitivity to land conditions in the Sahel: increases in albedo produce negative feedbacks on rainfall, and increases in soil moisture produce positive ones. The magnitude of these effects is, however, within a narrow range (Xue and Fennessy 2002). Desertification and deforestation amplified the precipitation deficit and the 1970–80 drought over the Sahel (Xue and Shukla 1993). On the other hand, Nicholson et al. (1998) found no progressive change in vegetation cover in the Sahel or the Saharan boundary 1980–95. In this context it is obvious that further research is needed, and not only in the context of the GGW.
The present study adds to the GGW related research with an experiment employing large ensembles of global simulations with the Model for Prediction Across Scales (MPAS). There are two scientific aims of the present study. The first is to evaluate the ability of the MPAS to reproduce the observed characteristics of the summer precipitation over the Sahel region. Such an evaluation is needed, as there are only few MPAS experiments with a focus on the African climate. Heinzeller et al. (2016) showed that MPAS can reproduce the atmospheric dynamics on both global and local scales. In their 11-month experiment, they identified a precipitation excess for the West African region. Maoyi and Abiodun (2021) found shortcomings of the MPAS in reproducing the pattern of precipitation related to positive and negative modes of the Botswana high. The second aim is to investigate the influence of a hypothetical GGW on Sahelian precipitation.
The present study is structured as follows: section 2 describes the applied model, reference data, investigation areas, evaluation indices, as well as the constructed GGW scenarios. The results obtained from the evaluation of the model and from the scenario runs are presented and discussed in section 3, and conclusions are drawn in section 4.
2. Material and methods
a. Applied model
The applied meteorological model is MPAS, version 7. MPAS is a modeling framework that includes an atmospheric model, an ocean model, and a land-ice model based on unstructured Voronoi meshes and C-grid discretization (Thuburn et al. 2009; Ringler et al. 2010). MPAS-atmosphere (Skamarock et al. 2012), used in the present study, is a global, fully compressible nonhydrostatic model (Klemp 2011). Heinzeller et al. (2016) provide a detailed evaluation of MPAS with an emphasis on technical aspects, including scalability on different high-performance computing systems. MPAS is a global model; therefore, the adverse effects of a nesting approach are avoided.
The static input fields used are the Moderate Resolution Imaging Spectroradiometer (MODIS) 20-class land cover and Global Multi-Resolution Terrain Elevation Data (GMTED2010) (Danielson and Gesch 2011) topography. The MODIS data are based on global land-cover climatology derived from MODIS data collected in 2001–10 at 500-m resolution (Broxton et al. 2014). MPAS can also utilize USGS land-use data based on the Global Land Cover Characterization (GLCC) data derived for the period 1992–93 at 1-km resolution (Loveland et al. 2000).
MPAS applies the Community Noah Land Surface Model (Noah-LSM) (Chen et al. 1996). Noah receives near-surface atmospheric forcing and simulates soil moisture and soil temperature, skin temperature, snowpack characteristics, canopy water content, surface water balance, and energy flux and water flux terms of the surface energy balance (Mitchell 2005). For vegetation parameters, such as leaf area index (LAI), maximum LAI, roughness length, root depth and emissivity fixed seasonal and from maximum and minimum values interpolated values associated with the land-cover categories are applied. Albedo and vegetation fraction is derived from monthly MODIS climatology (Friedl et al. 2002).
Numerous studies, mostly within the Weather Research and Forecasting (WRF) Model, have demonstrated appropriate representation of the involved processes, and reasonable sensitivity to changes in the diverse input parameters in various environments (Pathirana et al. 2014; McNally et al. 2015; Arnault et al. 2016). Glotfelty et al. (2021) stated that Noah-LSM is the most applied LSM within WRF-based studies in Africa. Obtained results range from superior performance (Igri et al. 2018) to rather limited performance in creating a climate change signal from land-use and land-cover changes (LULCC) in Africa (Noble et al. 2017). In simulations for China, Cao et al. (2015) showed that Noah-LSM performed better with adjusted parameters such as vegetation fraction, LAI, albedo, or emissivity.
The RRTMG (Clough et al. 2005) longwave and shortwave radiation scheme uses a fixed value for carbon dioxide, reflecting the conditions of the years around 2004. All performed simulations apply 55 vertical levels up to a height of 30 km, along with 4 soil levels.
The simulations in the present study use three mesh resolutions: The first resolution is approximately 60 km, with 163 842 cells. The second resolution is roughly 30 km, with 655 362 cells. Runs with those meshes use the mesoscale reference physics suite. The third resolution is a variable-resolution mesh with 835 586 cells. With 60–3 km, it extends the resolution in West Africa (center: 0° and 14°N) from 60 to 3 km. MPAS runs with this mesh use the scale-aware convection-permitting physics suite. Table 1 shows the chosen parameterizations schemes associated with the mesoscale and the convection-permitting physics suites of the standard model configuration.
Parameterization schemes used by the simulations.
b. Observational reference and investigation areas
The present investigation uses the Climate Research Unit (CRU) cru_ts4.04 monthly precipitation climatology (Harris et al. 2020) at 0.5° resolution and daily Climate Hazards Group InfraRed Precipitation with Station (CHIRPS) data (Funk et al. 2014) as well as NOAA Climate Data Record (CDR) of CPC Morphing Technique (CMORPH; https://doi.org/10.25921/w9va-q159) precipitation data, both at 0.25° resolution.
The results of the scenario simulations are analyzed in a large investigation area covering the area between −18°W and 40°E and between 4° and 24°N, referred to as GGW, and in three smaller areas. Two areas, referred to as western Sahel (SAH_W) and eastern Sahel (SAH_E) (Fig. 1), allow putting the results in the context of previous investigations of observed and simulated African precipitation by Dosio et al. (2021b,a). The third area, the central Sahel (SAC), covers the region where the variable MPAS mesh reaches 3-km resolution.
Investigation areas and 1200 UTC 1 Aug temperature (°C) in the simulation with 60-km resolution together with the MPAS 60-km mesh.
Citation: Earth Interactions 27, 1; 10.1175/EI-D-22-0013.1
Investigation areas and 1200 UTC 1 Aug temperature (°C) in the simulation with 60-km resolution together with the MPAS 60-km mesh.
Citation: Earth Interactions 27, 1; 10.1175/EI-D-22-0013.1
Investigation areas and 1200 UTC 1 Aug temperature (°C) in the simulation with 60-km resolution together with the MPAS 60-km mesh.
Citation: Earth Interactions 27, 1; 10.1175/EI-D-22-0013.1
c. Investigated indices of precipitation and further climate variables
Beside the mean seasonal precipitation rain rate (RR), the investigation applies several climate indices as defined by the Expert Team on Climate Change Detection and Indices (ETCCDI; Karl et al. 1999) (Table 2). In particular, these are as follows: number of wet days (RR1), intensity on wet days (INT), maximum daily precipitation (RX1day), number of consecutive wet days (CWD) and consecutive dry days (CDD), and number of days with heavy precipitation over 20 mm (here RR20), as used by Dosio et al. (2021b,a). They are calculated for land points only. Statistical significance has been tested applying Student’s t test. The required equality of variances has been tested using Levene’s test and the normality has been verified using the Shapiro–Wilk test. False discovery rate Wilks (2016) correction has been applied to field significance tests in spatial maps showing temperature and precipitation changes. In the false discovery rate, a new threshold p value is calculated based on the distribution of the p values of the analyzed field.
Precipitation indices used in this study.
In addition, several climate variables potentially influenced by the GGW are investigated. These are daily maximum temperature (TX) and its seasonal [June–September (JJAS)] maximum (TXx), daily minimum temperature (TN) and its seasonal minimum (TNx), upward heat flux at the surface (HFX), latent heat flux at the surface (LH), net surface shortwave radiation flux (GSW), surface runoff, and albedo.
d. Investigated scenarios
As the exact outline of the GGW belt is not available and several countries extend it by implementing additional action plans, only a hypothetical extent can be considered here. The following study delineates three potential courses within the 100–400-mm isohyets, as in the GGW plans, derived from CRU climatology 1971–2019. The employed precipitation ranges, the considered area, and the acronyms are shown in Table 3.
Land-cover change to category “woody savannah” within the isohyets specified from CRU climatology in the applied scenarios.
Available climate models use only a limited number of land-cover categories. Only a few of them can be found in the considered area. The use of a land-cover category tailored to the GGW is not yet possible, and thus any assignment to the MODIS scheme might be questioned. In the present study, in each considered band, the land-cover category has been changed to “woody savannah.” Figure 2 shows the standard S0 land use and changes for the S1, S2, and S3 scenarios. The area of the investigated scenarios ranges from 0.8 to 1.4 million km2. The total intervention area of the GGW effort is 1.56 million km2 (United Nations Convention to Combat Desertification 2020). However, this also covers the areas of additional national plans (Goffner et al. 2019). Changes in the land-cover category lead to changes in associated parameters. The maximum leaf area index (LAImax) increases for example within the GGW, depending on the replaced land-use category from 0.76 to 2.66 to the value of 3.66.
MODIS land-cover data at 30-km resolution (a) as implemented in MPAS 7.0 and run S0 and changed according to the different scenario settings (b) S1, (c) S2, and (d) S3. Only land-cover categories present in the area are shown: 1—evergreen needleleaf forest, 2—evergreen broadleaf forest, 4—deciduous broadleaf forest, 5—mixed forests, 6—closed shrubland, 7—open shrublands, 8—woody savannas, 9—savannas, 10—grasslands, 11—permanent wetlands, 12—croplands, 13—urban and built-up, 14—cropland/natural vegetation mosaic, 16—barren, and 17—water.
Citation: Earth Interactions 27, 1; 10.1175/EI-D-22-0013.1
MODIS land-cover data at 30-km resolution (a) as implemented in MPAS 7.0 and run S0 and changed according to the different scenario settings (b) S1, (c) S2, and (d) S3. Only land-cover categories present in the area are shown: 1—evergreen needleleaf forest, 2—evergreen broadleaf forest, 4—deciduous broadleaf forest, 5—mixed forests, 6—closed shrubland, 7—open shrublands, 8—woody savannas, 9—savannas, 10—grasslands, 11—permanent wetlands, 12—croplands, 13—urban and built-up, 14—cropland/natural vegetation mosaic, 16—barren, and 17—water.
Citation: Earth Interactions 27, 1; 10.1175/EI-D-22-0013.1
MODIS land-cover data at 30-km resolution (a) as implemented in MPAS 7.0 and run S0 and changed according to the different scenario settings (b) S1, (c) S2, and (d) S3. Only land-cover categories present in the area are shown: 1—evergreen needleleaf forest, 2—evergreen broadleaf forest, 4—deciduous broadleaf forest, 5—mixed forests, 6—closed shrubland, 7—open shrublands, 8—woody savannas, 9—savannas, 10—grasslands, 11—permanent wetlands, 12—croplands, 13—urban and built-up, 14—cropland/natural vegetation mosaic, 16—barren, and 17—water.
Citation: Earth Interactions 27, 1; 10.1175/EI-D-22-0013.1
This study does not investigate the question of whether such a change in the land cover is feasible in this area, considering the natural climatic and fertility conditions. Planting and growing trees in a vulnerable dry environment is not simple, even when local species are planted. Ba (2010) lists 30 woody and 22 herbaceous plants resilient to arid and semiarid climatic conditions proposed for cultivation in the GGW. These species have been selected taking into account socioeconomic values to the population and important ecological functions. With regard to aridity Elagib et al. (2021) conclude that 25% of the GGW area is not suitable for sustainable planting under rainfed conditions, and that around 93.0% of this area lies east of 10° longitude. On the other hand, Anchang et al. (2019) concluded that woody vegetation recovery is responsible for the postdrought greening of the Sahel and Sudanian regions.
e. Performed simulations
The global MPAS simulations use the SST and extent of the sea ice, both from ERA-Interim reanalysis data, as the only boundary condition. As a consequence, a global MPAS simulation does not necessarily simulate the observed period of a particular year. Each simulation yields weather patterns that fit to the SST and extent of sea ice used, and they are only one possible realization. Due to the nonlinearity of the climate system, it depends on the initial conditions. Therefore, an ensemble approach with several simulations has been applied. For each scenario, large ensembles of MPAS simulations are constructed. They consist of simulations made with different model resolutions, which can to some extent represent uncertainties due to limitations of the model. They are forced with several different boundary conditions representing the uncertainty due to forcing. Last, they apply a variety of different initial conditions, representing internal variability. Such ensembles can be used for comparison with the observations so as to evaluate the ability of the MPAS to simulate the observed climate and predict potential changes due to the GGW.
The specific years were chosen taking into account the spread in anomalies of the SST in the Gulf of Guinea (Fig. 3) assuming its central influence on the Sahelian precipitation (Son and Seo 2020). The considered period, highlighted in gray, is from 1997 to 2012. It was chosen around the year 2004 as the reference year of the RRTMG scheme. Particular years reflect years with warm, average, or cool SST in the area.
Anomaly of the mean SST in the area of the Gulf of Guinea as in the ERA-Interim data for the summer (JJA) season.
Citation: Earth Interactions 27, 1; 10.1175/EI-D-22-0013.1
Anomaly of the mean SST in the area of the Gulf of Guinea as in the ERA-Interim data for the summer (JJA) season.
Citation: Earth Interactions 27, 1; 10.1175/EI-D-22-0013.1
Anomaly of the mean SST in the area of the Gulf of Guinea as in the ERA-Interim data for the summer (JJA) season.
Citation: Earth Interactions 27, 1; 10.1175/EI-D-22-0013.1
MPAS simulations at 30-km resolution use boundary conditions with the SST and sea ice of the years 1998, 2005, 2010, and 2011 (Table 4) as in the ERA-Interim data; for each scenario and each year, five runs were initialized at 15–19 May. A total of 16 simulations at 60-km resolution use the SST of the years 1997–2012 initialized at 15 May (one for each SST year), and, in addition, four simulations with SST of the years 1998, 2005, 2010, and 2011 initialized at 19 May were done (one for each year of SST). Simulations with the largest extent of the GGW as in scenario S3 were repeated with adjusted albedo and vegetation fraction following conclusions of Glotfelty et al. (2021) who identified the satellite derived albedo climatology as source of additional errors in LULCC experiments. The monthly MODIS albedo and vegetation fraction were corrected by subtracting monthly constant values for albedo between 0.045 for May and 0.032 for September and adding values between 0.12 and 0.1 for the vegetation fraction. They have been derived as difference between the average albedo–vegetation fraction over the hypothetical GGW and the average over the land-cover category “woody savannah” present within the GGW investigation area. Figure 4 shows exemplarily the applied albedo and vegetation fraction changes for the month of July as in the repeated run of the scenario S3. The albedo variability within the S3 GGW course is comparatively small, and therefore the changes in albedo are also small. On the other hand, the change in the vegetation fraction is clearly visible. All simulations were run until 1 October.
Performed simulations.
Standard July MODIS (a) albedo and (b) vegetation fraction as in simulation S0 and (c) adjusted albedo and (d) adjusted vegetation fraction as in the additional scenario simulation S3.
Citation: Earth Interactions 27, 1; 10.1175/EI-D-22-0013.1
Standard July MODIS (a) albedo and (b) vegetation fraction as in simulation S0 and (c) adjusted albedo and (d) adjusted vegetation fraction as in the additional scenario simulation S3.
Citation: Earth Interactions 27, 1; 10.1175/EI-D-22-0013.1
Standard July MODIS (a) albedo and (b) vegetation fraction as in simulation S0 and (c) adjusted albedo and (d) adjusted vegetation fraction as in the additional scenario simulation S3.
Citation: Earth Interactions 27, 1; 10.1175/EI-D-22-0013.1
Because of the large demand on computational resources, variable-mesh simulations at 60–3-km resolution were performed with SST of the years 2010 and 2011 and the month of August only (MP60-3). These simulations were initialized on 23–27 June. The available simulations were regridded to 0.25° resolution of the CHIPRS reference and lumped to standard and scenario ensembles with 40 members for the JJAS season and 50 members for the month of August, referred to as MPAS and MPASa, respectively. Bilinear interpolation was applied as regridding procedure. These ensembles are the subject of the present investigation.
3. Results
a. Evaluation
Available precipitation climatologies for the African continent reveal substantial differences, especially in the area of the Gulf of Guinea and the Sahel (Dosio et al. 2021a). Sylla et al. (2013) concluded that systematic differences in mean precipitation are particularly visible in related daily precipitation statistics, such as extremes, dry spells, or frequency of wet days. Masunaga et al. (2019) attributed the deficiencies primarily to low gauge density in some areas. On the other hand, simulation of the precipitation and its characteristics in both wet and semiarid regions is still a big challenge. General circulation models (GCM) as well as regional climate models (RCM) reveal substantial limitations for correctly representing the physical process driving the African climate. In consequence, GCM and RCM simulations of the past and future African precipitation have large discrepancies (Dosio et al. 2021a). Nikiema et al. (2017) found dry biases in GCM/RCM simulations up to 60% over the Sahel region.
Evaluation of the MPAS simulations is performed in comparison with 16-yr periods of the CHIRPS (1997–2012) and CMORPH (1998–2013) references for the precipitation of the month of August and the summer season [June–August (JJA)]. JJA has been chosen to allow a comparison with the results obtained by Dosio et al. (2021b,a), who analyzed 20 precipitation climatologies and several ensembles of various GCM/RCM simulations. There, however, the considered time period was 1980–2010. Because MPAS simulations with the applied land-use data and the RRTMG scheme rather fit do the years around 2004 until 2010, an exact specification of the comparison period is hardly possible, and some variation has to be accepted here.
Relative to the CHIRPS 1997–2012 reference, the MPAS simulation underestimates the summer (JJA) precipitation of the GGW area by 0.9 mm day−1, which is roughly −25%. In relation to CMORPH, the dry bias is −0.1 mm day−1 (−4%) smaller. Table 5 provides details for the area mean values of the investigated indices. However, Figs. 5a, 5c, and 5e reveal considerable regional biases. In comparison with the observational reference there is an overestimation by about 33% in area SAH_E, an underestimation by about 41% is visible in area SAH_W. For SAH_E, the mean summer precipitation is in the range of the observational reference compiled by Dosio et al. (2021a). For SAH_W this is not the case. The MPAS simulations reveal smaller biases in RX1day and RR1 but there is a dry bias in precipitation intensity on wet days of about 22%–48%.
Mean values of precipitation indices over the investigation areas for the JJA season, both observed and simulated. MPAS is the mean value of the standard S0 ensemble. Minimum, mean, and maximum values of observational datasets for satellite (label S) and gauge (label G) products as presented by Dosio et al. (2021a) are respectively given from left to right in each column.
Observed mean precipitation as (a),(b) average of CHIRPS 1997–2012 and CMORPH 1998–2013, (c),(d) simulated MPAS, and (e),(f) difference MPAS − average of CHIRPS and CMORPH for (left) the summer (JJA) season (right) August.
Citation: Earth Interactions 27, 1; 10.1175/EI-D-22-0013.1
Observed mean precipitation as (a),(b) average of CHIRPS 1997–2012 and CMORPH 1998–2013, (c),(d) simulated MPAS, and (e),(f) difference MPAS − average of CHIRPS and CMORPH for (left) the summer (JJA) season (right) August.
Citation: Earth Interactions 27, 1; 10.1175/EI-D-22-0013.1
Observed mean precipitation as (a),(b) average of CHIRPS 1997–2012 and CMORPH 1998–2013, (c),(d) simulated MPAS, and (e),(f) difference MPAS − average of CHIRPS and CMORPH for (left) the summer (JJA) season (right) August.
Citation: Earth Interactions 27, 1; 10.1175/EI-D-22-0013.1
Results obtained for August, the month with the precipitation peak in the region, are similar (Table 6 and Figs. 5b,d,f). In comparison with the observational reference, MPAS underestimates the area mean precipitation in GGW by 14%, in SAC by 27%, and in SAH_W by 30%. In SAH_E there is an overestimation by about 19%. Again, while RR, RR1, and RX1day are reasonably simulated in all investigation areas, the major shortcoming is in the precipitation intensity.
Mean values of precipitation indices over the investigation areas for August, both observed and simulated. MPASa is the mean value of the standard S0 ensemble with 50 members.
Dosio et al. (2021b) investigated the performance of CMIP5 and CMIP6 global models as well as the Coordinated Regional Climate Downscaling Experiment (CORDEX) and Coordinated Output for Regional Evaluations (CORE) regional modeling experiments in simulating African daily precipitation. They conclude that there is a large spread between the models and a general underestimation of the precipitation peak for the Sahel region. In comparison, MPAS simulations are in the range of the models investigated there.
It is worth pointing out that there is only a small difference between the results obtained with the 30- and 60-km resolutions. Pilon et al. (2016) evaluated the parameterization of shallow cumulus microphysics, boundary layer, and deep cumulus, concluding that there was only little difference in the fidelity of the model for grid spacings between 60 and 15 km. In the present investigation, the results obtained with 30- and 60-km resolutions of the MPAS model are similar in all statistics and reveal for example in RR only small differences of about 0.2 mm day−1.
Convection permitting simulations with variable mesh seem to reduce the dry bias over SAH_W. But this area is to a large extent already in the transition zone. In SAC, which is within the 3-km mesh, the dry bias is not smaller. There is, however, an interesting difference: MPAS simulations show in SAC only a small bias in INT, but they underestimate RR1.
Various reasons for the obvious shortcomings in the simulation of the Sahelian precipitation by the RCM models have been discussed. Panitz et al. (2014) explain underestimation of the rainfall peak in regions affected by the passage of the monsoon by misplacement of the center of the monsoon and the underestimation of its intensity and to the northern shift of the West African heat low. In their experiment with the COSMO in Climate Mode (CCLM) model, Dieng et al. (2017) found that the model simulated the monsoon precipitation belt too far north, which resulted in an overestimation of precipitation in the Sahel zone by up to 60%. MPAS correctly simulates the northern extent of the precipitation belt, but it fails in transporting enough moisture from the Atlantic. From a modeling study in West Africa, Arnault et al. (2021) identified lateral terrestrial water flow and its contribution to land surface evaporation as a source of uncertainty in modeled precipitation. Diallo et al. (2017) state that the underestimation of precipitation intensity may also be related to the underestimation of the surface shortwave radiation and latent heat flux.
b. Scenario runs
Analysis of the scenario runs is performed for the rainy season JJAS and August as the month with the maximum precipitation. The investigated precipitation indices are RR, RR1, RR20, CDW, and CDD.
The standard MPAS ensemble reveals in the GGW an area mean JJAS precipitation of 2.77 mm day−1. The S1, S2 and S3 scenarios show only small differences in the range from −0.03 to 0.01 mm day−1. Student’s t test was computed to determine whether the obtained differences are significant at the 5% significance level. There are no extreme outliers in the data. The required equality of variances is fulfilled, as assessed by Levene’s test. A Shapiro–Wilk test finds that the distributions of S0, S1, S2, and S3 are normally distributed. The obtained differences are in all cases not significant, with p values ranging from 0.434 to 0.828.
The same results apply to the investigation areas SAH_W, SAH_E, and SAC, with a slightly increased difference of 0.13 mm day−1 in SAC. In the arid environment with xerophyte vegetation the prescribed land-use change cannot provide additional moisture influencing the precipitations.
Sahelian summer precipitation shows pronounced interannual and decadal variations. Rodríguez-Fonseca et al. (2015) attributed large parts of this variability to slowly varying climate subcomponents, such as SST over the tropical Atlantic, the Mediterranean, and the Pacific Ocean, and land surface conditions. Abiodun et al. (2008) showed that desertification increases the meridional temperature gradient, which increases the monsoon flow over the Guinean region. As a result, rainfall is reduced over the desertification region and increased in the southern part. Deforestation decreases the surface friction experienced by the flow over the Guinean region. This reduces precipitation over the entire West African region. Such changes in land surface conditions have involved large areas of the Sahara–Sahelian zone. In comparison with that, the area of the simulated green wall is small and therefore probably unable to influence the total amount of precipitation, neither through increased evaporation nor inducing changes in physical processes driving the Sahelian precipitation. The performed MPAS simulations support this theory.
Figure 6 shows boxplots of precipitation indices RR1, RR20, CWD, and CDD simulated by the standard and scenario runs for the JJAS season. Again, notable changes are not present in either the mean values or in the distribution pattern. Also, there is only small variation visible between the scenario ensembles.
Boxplot of precipitation indices RR1, RR20 (here labeled R20), CDD, and CWD simulated by standard S0 and scenario runs S1, S2 and S3 for the JJAS season for investigation areas (top) GGW, (middle) SAH_W, and (bottom) SAH_E. Red dots indicate mean values, and black open circles are outliers.
Citation: Earth Interactions 27, 1; 10.1175/EI-D-22-0013.1
Boxplot of precipitation indices RR1, RR20 (here labeled R20), CDD, and CWD simulated by standard S0 and scenario runs S1, S2 and S3 for the JJAS season for investigation areas (top) GGW, (middle) SAH_W, and (bottom) SAH_E. Red dots indicate mean values, and black open circles are outliers.
Citation: Earth Interactions 27, 1; 10.1175/EI-D-22-0013.1
Boxplot of precipitation indices RR1, RR20 (here labeled R20), CDD, and CWD simulated by standard S0 and scenario runs S1, S2 and S3 for the JJAS season for investigation areas (top) GGW, (middle) SAH_W, and (bottom) SAH_E. Red dots indicate mean values, and black open circles are outliers.
Citation: Earth Interactions 27, 1; 10.1175/EI-D-22-0013.1
As compared with the standard run, scenario simulations introduce regional variations. Figure 7 shows maps of the differences between the wet-days index RR1 (scenario − S0) for the JJAS season and the GGW area. Positive and negative changes by up to ±6 days are present. The sign of the changes varies between the scenarios and only in the area of Burkina Faso and parts of Mali is there an agreement on RR1 reduction and in the Guinea Coast area on an increase. However, only in a few grid cells in the GGW area are such differences significant at the 5% level. Thus, no reliable attribution of the simulated changes to the prescribed land-use change is possible.
Change in JJAS wet-days RR1 over the target area for the different land-use scenarios (a) S1, (b) S2, and (c) S3 relative to the standard run S0. Statistical significance at 5% level as indicated by black dots is present only without false discovery rate correction.
Citation: Earth Interactions 27, 1; 10.1175/EI-D-22-0013.1
Change in JJAS wet-days RR1 over the target area for the different land-use scenarios (a) S1, (b) S2, and (c) S3 relative to the standard run S0. Statistical significance at 5% level as indicated by black dots is present only without false discovery rate correction.
Citation: Earth Interactions 27, 1; 10.1175/EI-D-22-0013.1
Change in JJAS wet-days RR1 over the target area for the different land-use scenarios (a) S1, (b) S2, and (c) S3 relative to the standard run S0. Statistical significance at 5% level as indicated by black dots is present only without false discovery rate correction.
Citation: Earth Interactions 27, 1; 10.1175/EI-D-22-0013.1
The picture of simulated changes in August is similar. Figure 8 depicts boxplots derived from the MPASa ensembles with 50 members for the SAC investigation area revealing only small changes in the mean values and in the distribution. Simulated regional changes within the GGW area show in all scenarios throughout the Sahel a small reduction of the number of wet days. As for the JJAS season, regional positive and negative changes are visible within the GGW area. However, only a few cells reveal statistically significant differences (Fig. 9).
Boxplot of precipitation indices RR1, RR20 (here labeled R20), CDD, and CWD simulated by standard S0 and scenario runs S1, S2 and S3 for the month of August and the SAC investigation area. Red dots indicate mean values, and black open circles are outliers.
Citation: Earth Interactions 27, 1; 10.1175/EI-D-22-0013.1
Boxplot of precipitation indices RR1, RR20 (here labeled R20), CDD, and CWD simulated by standard S0 and scenario runs S1, S2 and S3 for the month of August and the SAC investigation area. Red dots indicate mean values, and black open circles are outliers.
Citation: Earth Interactions 27, 1; 10.1175/EI-D-22-0013.1
Boxplot of precipitation indices RR1, RR20 (here labeled R20), CDD, and CWD simulated by standard S0 and scenario runs S1, S2 and S3 for the month of August and the SAC investigation area. Red dots indicate mean values, and black open circles are outliers.
Citation: Earth Interactions 27, 1; 10.1175/EI-D-22-0013.1
Change in August wet-days RR1 over the target area for the different land-use scenarios (a) S1, (b) S2, and (c) S3 relative to the standard run S0. Statistical significance at 5% level as indicated by black dots is present only without false discovery rate correction.
Citation: Earth Interactions 27, 1; 10.1175/EI-D-22-0013.1
Change in August wet-days RR1 over the target area for the different land-use scenarios (a) S1, (b) S2, and (c) S3 relative to the standard run S0. Statistical significance at 5% level as indicated by black dots is present only without false discovery rate correction.
Citation: Earth Interactions 27, 1; 10.1175/EI-D-22-0013.1
Change in August wet-days RR1 over the target area for the different land-use scenarios (a) S1, (b) S2, and (c) S3 relative to the standard run S0. Statistical significance at 5% level as indicated by black dots is present only without false discovery rate correction.
Citation: Earth Interactions 27, 1; 10.1175/EI-D-22-0013.1
The performed investigations reveal no changes in the northern extent of the rain belt. Such changes can possibly result from an increased roughness length’s influencing the monsoon flow over the region (Oguntunde et al. 2014).
Table 7 shows the area mean values for temperature, radiation fluxes and surface runoff as in simulations performed with 60-km resolution for the investigated areas SAH_W, SAH_E, and SAC. Obtained results indicate, similar to precipitation, only small changes, none of which are significant at the 5% significance level.
Simulated seasonal (JJAS) area mean values of selected variables as in the simulations with 60-km resolution. See section 2c for the definitions of the variables.
Contrary to the results obtained by Saley et al. (2019), the present study does not provide any significant evidence for GGW-induced changes in the characteristics of the summer precipitation: neither within the Sahel nor in the neighboring regions. Despite relatively large changes in the land cover, related changes in the investigated indices are comparatively small, which raises the question about possible reasons. MPAS runs at 60-km resolution with the S3 scenario applying the largest extent of the hypothetical GGW with adjusted albedo and vegetation fraction show small changes in the investigated indices in comparison with the standard run S0, all of which are not significant and are similar to changes found between the investigated scenarios. Thus, applied climatological MODIS input data do not weaken the applied investigation design.
On the other hand, potential impact cannot entirely be neglected. Figure 10 shows the changes in the JJAS mean precipitation RR in the GGW area in the scenario runs S3 without and with adjusted monthly albedo relative to the standard run S0. Majority of the grid cells reveal only not significant changes, however, there is change in the sign of the area mean value from −0.03 to +0.03 mm day−1. Albedo impact was demonstrated by Charney (1975) who argued that increase in albedo resulting from a reduction of vegetation in the Sahara–Sahelian region causes sinking motion and additional drying perpetuating the arid conditions. The certain role of vegetation–albedo feedback is also demonstrated by paleoclimate simulations. Singh et al. (2022) showed that inclusion of woody forest on higher latitudes and shrubs and steppes over northern Africa played a central role in the additional monsoon season rainfall. Pausata et al. (2016) stated that reproduction of proxy records was only possible with inclusion of Saharan vegetation and dust changes, both influencing the albedo.
Change in JJAS precipitation RR over the target area relative to the standard run S0 in the (a) 60-km-resolution S3 scenario run and (b) 60-km-resolution S3a scenario run with adjusted monthly albedo. Statistical significance at 5% level as indicated by black dots is present only without false discovery rate correction.
Citation: Earth Interactions 27, 1; 10.1175/EI-D-22-0013.1
Change in JJAS precipitation RR over the target area relative to the standard run S0 in the (a) 60-km-resolution S3 scenario run and (b) 60-km-resolution S3a scenario run with adjusted monthly albedo. Statistical significance at 5% level as indicated by black dots is present only without false discovery rate correction.
Citation: Earth Interactions 27, 1; 10.1175/EI-D-22-0013.1
Change in JJAS precipitation RR over the target area relative to the standard run S0 in the (a) 60-km-resolution S3 scenario run and (b) 60-km-resolution S3a scenario run with adjusted monthly albedo. Statistical significance at 5% level as indicated by black dots is present only without false discovery rate correction.
Citation: Earth Interactions 27, 1; 10.1175/EI-D-22-0013.1
Studies simulating LULCC in sub-Saharan Africa applying the Noah-LSM and other LSMs as the land surface model in WRF draw different conclusions about the achieved performance. Recently, Glotfelty et al. (2021) stated that LSMs reveal significant deficiencies when simulated impacts of land-use and land-cover changes in sub-Saharan Africa. The authors applied three versions of Noah and two versions of the Community Land Model (CLM) (Lawrence et al. 2019) with the WRF Model and found that due to overestimated albedo resulting from the MODIS data, Noah-LSM revealed slightly larger cold and dry biases than CLM. In general, they questioned the suitability of the standard LSMs Noah-LSM as well as CLM applicable with WRF in LULCC experiments in Africa. Hagos et al. (2014) argued that the weak signal in such experiments is caused by the model falling in erroneous moisture- or energy-limited regimes when the simulations reveal a dry or wet bias relative to observations and reanalyses.
Odoulami et al. (2019) detected an impact of Savannah afforestation, however, in a much larger area between 8° and 12°N from Atlantic until 20°E and for the future climate. Most significant changes were present over the afforestation area. WRF with Noah-LSM and RegCM4 (Giorgi et al. 2012) with Biosphere–Atmosphere Transfer Scheme (BATS; Dickinson et al. 1993) revealed a realistic reproduction of various precipitation indices, with significant biases. Obtained changes in latent heat and sensible heat fluxes are in a range comparable to results obtained in the present experiment.
Another source of possible shortcomings is the initialization of the soil properties in the performed short MPAS runs. Long time runs accumulating possible changes in the GGW area on the soil moisture might be required here.
Limiting the investigation area to the exact extent of the hypothetical GGW only, yields in the S1 scenario reduction of TX, TXx, TN, and TNx the JJAS mean in range between −0.3° and −0.1°C. Decrease in daily minimum temperature TN of −0.3°C is significant at the 95% significance level. In the S2 scenario, this reduction is on the order of −0.1°C. In the scenario S3 with the most southern GGW course, no changes are present.
4. Conclusions
Multiple global runs of the MPAS model with SST and sea ice from the ERA-Interim reanalysis as the only boundary condition have been performed with a focus on the Sahel zone for the rainy season from June to September, and with August as the month with the peak precipitation. Using three different model resolutions and two different configurations addresses the uncertainty due to model limitations. Using initializations in different years and on different days addresses the uncertainty due to boundary conditions and the internal variability. The obtained data ensembles have been applied to study the influence of a hypothetical course of the planned Great Green Wall on the precipitation in the Sahel region.
Evaluation of the results obtained with the standard MPAS runs against the observational reference shows for the GGW investigation area covering the Sahel zone an underestimation of the JJAS area mean precipitation by roughly 14%. However, the mean value has a wet bias by about 33% in the eastern part and a stronger dry bias of 41% in the western part.
The present study does not provide any significant evidence for GGW-induced changes in the characteristics of the summer precipitation, for positive changes within the Sahel supporting the forestation activities, or for potentially adverse changes in the neighboring regions. Also, changes simulated for other climate variables potentially influenced by the GGW as temperature, radiation fluxes or runoff are comparatively small. However, at least three sources of uncertainty have to be mentioned here. As in many sensitivity modeling studies, the present study applies a complete instantaneous regional land-cover change. Yu et al. (2017) argued that this has limited real-world relevance, as observed vegetation changes are typically heterogeneous in space and transient in time. Furthermore, SST driven simulations failed to explain the present regreening process. The latter hints at shortcomings in the formulation of the model, most probably in the land–atmosphere interaction and to the second source of uncertainty. As with desertification (Nicholson et al. 1998), the true extent of reforestation and its relationship to meteorological factors can only be evaluated by recognizing its complexity. Additional uncertainty can be related to the applied albedo climatology.
The study considered the present time precipitation only. Monerie et al. (2021) stated that climate change will drive major perturbations of the West African summer monsoon, resulting in an increase in precipitation over the central Sahel and a decrease in precipitation over the western Sahel. Sylla et al. (2018) concluded that under the 2°C scenario, there will be severe freshwater shortages and a general decline of the basin-scale irrigation potential in West Africa. Such changes will add to the complexity of simulating changes in the Sahelian precipitation related to changes in the land cover. Without any doubt, the pan-African Great Green Wall effort has a large potential for societal benefits in various areas and can help to mitigate the effects of climate change. Also, it can trigger projects aiming at improvements in the simulation of the Sahelian climate. However, a statistically significant impact on precipitation cannot be concluded in the efforts made in this study with MPAS.
Acknowledgments.
The authors gratefully acknowledge the Gauss Centre for Supercomputing e.V. (https://www.gauss-centre.eu) for funding this project by providing computing time on the GCS Supercomputer SuperMUC-NG at the Leibniz Supercomputing Centre (https://www.lrz.de).
Data availability statement.
The data generated for this study are available at the Radar4KIT repository (https://radar.kit.edu/radar/en/dataset/lEoqMATCdUtEIKmm?token=DprMIcyxaxYizyeDAtoT). Required calculations were performed and figures were created by applying the CDO (Schulzweida 2021), NCO (Zender 2022), R (R Core Team 2021), and NCL (NCAR 2021) software packages.
REFERENCES
Abiodun, B. J., J. S. Pal, E. A. Afiesimama, W. J. Gutowski, and A. Adedoyin, 2008: Simulation of West African monsoon using RegCM3. Part II: Impacts of deforestation and desertification. Theor. Appl. Climatol., 93, 245–261, https://doi.org/10.1007/s00704-007-0333-1.
Anchang, J. Y., and Coauthors, 2019: Trends in woody and herbaceous vegetation in the savannas of West Africa. Remote Sens., 11, 576, https://doi.org/10.3390/rs11050576.
Arnault, J., R. Knoche, J. Wei, and H. Kunstmann, 2016: Evaporation tagging and atmospheric water budget analysis with WRF: A regional precipitation recycling study for West Africa. Water Resour. Res., 52, 1544–1567, https://doi.org/10.1002/2015WR017704.
Arnault, J., and Coauthors, 2021: Lateral terrestrial water flow contribution to summer precipitation at continental scale—A comparison between Europe and West Africa with WRF-Hydro-tag ensembles. Hydrol. Processes, 35, e14183, https://doi.org/10.1002/hyp.14183.
Ba, A. T., 2010: Grande Muraille Verte: Choix des espèces végétales. Le Projet Majeur Africain de la Grande Muraille Verte, A. Dia and R. Duponnois, Eds., Institut de Recherche pour le Développent, 31–38.
Bamba, A., and Coauthors, 2019: Effect of the African greenbelt position on West African summer climate: A regional climate modeling study. Theor. Appl. Climatol., 137, 309–322, https://doi.org/10.1007/s00704-018-2589-z.
Broxton, P. D., X. Zeng, D. Sulla-Menashe, and P. A. Troch, 2014: A global land cover climatology using MODIS data. J. Appl. Meteor. Climatol., 53, 1593–1605, https://doi.org/10.1175/JAMC-D-13-0270.1.
Cao, Q., D. Yu, M. Georgescu, Z. Han, and J. Wu, 2015: Impacts of land use and land cover change on regional climate: A case study in the agro-pastoral transitional zone of China. Environ. Res. Lett., 10, 124025, https://doi.org/10.1088/1748-9326/10/12/124025.
Charney, J. G., 1975: Dynamics of deserts and drought in the Sahel. Quart. J. Roy. Meteor. Soc., 101, 193–202, https://doi.org/10.1002/qj.49710142802.
Chen, F., and Coauthors, 1996: Modeling of land-surface evaporation by four schemes and comparison with FIFE observations. J. Geophys. Res., 101, 7251–7268, https://doi.org/10.1029/95JD02165.
Clough, S. A., M. W. Shephard, E. J. Mlawer, J. S. Delamere, M. J. Iacono, K. Cady-Pereira, S. Boukabara, and P. D. Brown, 2005: Atmospheric radiative transfer modeling: A summary of the AER codes. J. Quant. Spectrosc. Radiat. Transfer, 91, 233–244, https://doi.org/10.1016/j.jqsrt.2004.05.058.
Danielson, J. J., and D. B. Gesch, 2011: Global multi-resolution terrain elevation data 2010 (GMTED2010). U.S. Geological Survey Open-File Rep. 2011-1073, 23 pp., https://doi.org/10.3133/ofr20111073.
Diallo, F. B., F. Hourdin, C. Rio, A.-K. Traore, L. Mellul, F. Guichard, and L. Kergoat, 2017: The surface energy budget computed at the grid-scale of a climate model challenged by station data in West Africa. J. Adv. Model. Earth Syst., 9, 2710–2738, https://doi.org/10.1002/2017MS001081.
Dickinson, R. E., A. Henderson-Sellers, and P. J. Kennedy, 1993: Biosphere–Atmosphere Transfer Scheme (BATS) version 1e as coupled to the NCAR community climate model. NCAR Tech. Note NCAR/TN-387+STR, 80 pp., https://doi.org/10.5065/D67W6959.
Dieng, D., G. Smiatek, J. Bliefernicht, D. Heinzeller, A. Sarr, A. T. Gaye, and H. Kunstmann, 2017: Evaluation of the COSMO-CLM high-resolution climate simulations over West Africa. J. Geophys. Res. Atmos., 122, 1437–1455, https://doi.org/10.1002/2016JD025457.
Dosio, A., I. Pinto, C. Lennard, M. B. Sylla, C. Jack, and G. Nikulin, 2021a: What can we know about recent past precipitation over Africa? Daily characteristics of African precipitation from a large ensemble of observational products for model evaluation. Earth Space Sci., 8, e2020EA001466, https://doi.org/10.1029/2020EA001466.
Dosio, A., and Coauthors, 2021b: Projected future daily characteristics of African precipitation based on global (CMIP5, CMIP6) and regional (CORDEX, CORDEX-CORE) climate models. Climate Dyn., 57, 3135–3158, https://doi.org/10.1007/s00382-021-05859-w.
Elagib, N. A., M. Khalifa, Z. Babker, A. A. Musa, and A. H. Fink, 2021: Demarcating the rainfed unproductive zones in the African Sahel and Great Green Wall regions. Land Degrad. Dev., 32, 1400–1411, https://doi.org/10.1002/ldr.3793.
Friedl, M. A., and Coauthors, 2002: Global land cover mapping from MODIS: Algorithms and early results. Remote Sens. Environ., 83, 287–302, https://doi.org/10.1016/S0034-4257(02)00078-0.
Funk, C. C., and Coauthors, 2014: A quasi-global precipitation time series for drought monitoring. U.S. Geological Survey Data Series Tech. Rep. 832, 12 pp., https://doi.org/10.3133/ds832.
Giorgi, F., and Coauthors, 2012: RegCM4: Model description and preliminary tests over multiple CORDEX domains. Climate Res., 52, 7–29, https://doi.org/10.3354/cr01018.
Glotfelty, T., D. Ramírez-Mejía, J. Bowden, A. Ghilardi, and J. J. West, 2021: Limitations of WRF land surface models for simulating land use and land cover change in sub-Saharan Africa and development of an improved model (CLM-AF v. 1.0). Geosci. Model Dev., 14, 3215–3249, https://doi.org/10.5194/gmd-14-3215-2021.
Goffner, D., H. Sinare, and L. J. Gordon, 2019: The Great Green Wall for the Sahara and the Sahel initiative as an opportunity to enhance resilience in Sahelian landscapes and livelihoods. Reg. Environ. Change, 19, 1417–1428, https://doi.org/10.1007/s10113-019-01481-z.
Hagos, S., L. R. Leung, Y. Xue, A. Boone, F. de Sales, N. Neupane, M. Huang, and J.-H. Yoon, 2014: Assessment of uncertainties in the response of the African monsoon precipitation to land use change simulated by a regional model. Climate Dyn., 43, 2765–2775, https://doi.org/10.1007/s00382-014-2092-x.
Harris, I., T. J. Osborn, P. Jones, and D. Lister, 2020: Version 4 of the CRU TS monthly high-resolution gridded multivariate climate dataset. Sci. Data, 7, 109, https://doi.org/10.1038/s41597-020-0453-3.
Heinzeller, D., M. G. Duda, and H. Kunstmann, 2016: Towards convection-resolving, global atmospheric simulations with the Model for Prediction Across Scales (MPAS) v3.1: An extreme scaling experiment. Geosci. Model Dev., 9, 77–110, https://doi.org/10.5194/gmd-9-77-2016.
Igri, P. M., R. S. Tanessong, D. A. Vondou, J. Panda, A. Garba, F. K. Mkankam, and A. Kamga, 2018: Assessing the performance of WRF model in predicting high-impact weather conditions over central and western Africa: An ensemble-based approach. Nat. Hazards, 93, 1565–1587, https://doi.org/10.1007/s11069-018-3368-y.
Karl, T. R., N. Nicholls, and A. Ghazi, 1999: CLIVAR/GCOS/WMO workshop on indices and indicators for climate extremes workshop summary. Weather and Climate Extremes, T. Karl, N. Nicholls, and A. Ghazi, Eds., Springer, 3–7, https://doi.org/10.1007/978-94-015-9265-9_2.
Klemp, J. B., 2011: A terrain-following coordinate with smoothed coordinate surfaces. Mon. Wea. Rev., 139, 2163–2169, https://doi.org/10.1175/MWR-D-10-05046.1.
Lawrence, D. M., and Coauthors, 2019: The Community Land Model version 5: Description of new features, benchmarking, and impact of forcing uncertainty. J. Adv. Model. Earth Syst., 11, 4245–4287, https://doi.org/10.1029/2018MS001583.
Loveland, T. R., B. C. Reed, J. F. Brown, D. O. Ohlen, Z. Zhu, L. Yang, and J. W. Merchant, 2000: Development of a global land cover characteristics database and IGBP DISCover from 1 km AVHRR data. Int. J. Remote Sens., 21, 1303–1330, https://doi.org/10.1080/014311600210191.
Maoyi, M. L., and B. J. Abiodun, 2021: How well does MPAS-atmosphere simulate the characteristics of the Botswana high? Climate Dyn., 57, 2109–2128, https://doi.org/10.1007/s00382-021-05797-7.
Masunaga, H., M. Schröder, F. A. Furuzawa, C. Kummerow, E. Rustemeier, and U. Schneider, 2019: Inter-product biases in global precipitation extremes. Environ. Res. Lett., 14, 125016, https://doi.org/10.1088/1748-9326/ab5da9.
McNally, A., and Coauthors, 2015: Calculating crop water requirement satisfaction in the West Africa Sahel with remotely sensed soil moisture. J. Hydrometeor., 16, 295–305, https://doi.org/10.1175/JHM-D-14-0049.1.
Mitchell, K., 2005: The Community Noah Land-Surface Model (LSM). NCAR User’s Guide public release version 2.7.1, 26 pp., https://ral.ucar.edu/sites/default/files/public/product-tool/unified-noah-lsm/Noah_LSM_USERGUIDE_2.7.1.pdf.
Monerie, P.-A., B. Pohl, and M. Gaetani, 2021: The fast response of Sahel precipitation to climate change allows effective mitigation action. npj Climate Atmos. Sci., 4, 24, https://doi.org/10.1038/s41612-021-00179-6.
NCAR, 2021: The NCAR Command Language (version 6.6.2). UCAR/NCAR/CISL/TDD, https://doi.org/10.5065/D6WD3XH5.
Nicholson, S. E., C. J. Tuckerand, and M. B. Ba, 1998: Desertification, drought, and surface vegetation: An example from the West African Sahel. Bull. Amer. Meteor. Soc., 79, 815–830, https://doi.org/10.1175/1520-0477(1998)079<0815:DDASVA>2.0.CO;2.
Nieto, R., L. Gimeno, and R. M. Trigo, 2006: A Lagrangian identification of major sources of Sahel moisture. Geophys. Res. Lett., 33, L18707, https://doi.org/10.1029/2006GL027232.
Nikiema, P. M., M. B. Sylla, K. Ogunjobi, I. Kebe, P. Gibba, and F. Giorgi, 2017: Multi-model CMIP5 and CORDEX simulations of historical summer temperature and precipitation variabilities over West Africa. Int. J. Climatol., 37, 2438–2450, https://doi.org/10.1002/joc.4856.
Noble, E., L. M. Druyan, and M. Fulakeza, 2017: The sensitivity of WRF daily summertime simulations over West Africa to alternative parameterizations. Part II: Precipitation. Mon. Wea. Rev., 145, 215–233, https://doi.org/10.1175/MWR-D-15-0294.1.
Odoulami, R. C., B. J. Abiodun, and A. E. Ajayi, 2019: Modelling the potential impacts of afforestation on extreme precipitation over West Africa. Climate Dyn., 52, 2185–2198, https://doi.org/10.1007/s00382-018-4248-6.
Oguntunde, P. G., B. J. Abiodun, G. Lischeid, and C. Merz, 2014: Modelling the impacts of reforestation on the projected hydroclimatology of Niger River Basin, West Africa. Ecohydrology, 7, 163–176, https://doi.org/10.1002/eco.1343.
Panitz, H.-J., A. Dosio, M. Büchner, D. Lüthi, and K. Keuler, 2014: COSMO-CLM (CCLM) climate simulations over CORDEX-Africa domain: Analysis of the ERA-Interim driven simulations at 0.44° and 0.22° resolution. Climate Dyn., 42, 3015–3038, https://doi.org/10.1007/s00382-013-1834-5.
Pathirana, A., H. B. Denekew, W. Veerbeek, C. Zevenbergen, and A. T. Banda, 2014: Impact of urban growth-driven landuse change on microclimate and extreme precipitation—A sensitivity study. Atmos. Res., 138, 59–72, https://doi.org/10.1016/j.atmosres.2013.10.005.
Pausata, F. S. R., G. Messori, and Q. Zhang, 2016: Impacts of dust reduction on the northward expansion of the African monsoon during the Green Sahara period. Earth Planet. Sci. Lett., 434, 298–307, https://doi.org/10.1016/j.epsl.2015.11.049.
Pilon, R., C. Zhang, and J. Dudhia, 2016: Roles of deep and shallow convection and microphysics in the MJO simulated by the model for prediction across scales. J. Geophys. Res. Atmos., 121, 10 575–10 600, https://doi.org/10.1002/2015JD024697.
R Core Team, 2021: R: A language and environment for statistical computing. R Project, https://www.R-project.org/.
Ringler, T. D., J. Thuburn, J. B. Klemp, and W. C. Skamarock, 2010: A unified approach to energy conservation and potential vorticity dynamics for arbitrarily-structured C-grids. J. Comput. Phys., 229, 3065–3090, https://doi.org/10.1016/j.jcp.2009.12.007.
Rodríguez-Fonseca, B., and Coauthors, 2015: Variability and predictability of West African droughts: A review on the role of sea surface temperature anomalies. J. Climate, 28, 4034–4060, https://doi.org/10.1175/JCLI-D-14-00130.1.
Saley, I. A., S. Salack, I. S. Sanda, M. S. Moussa, A. L. Bonkaney, M. Ly, and M. Fodé, 2019: The possible role of the Sahel greenbelt on the occurrence of climate extremes over the West African Sahel. Atmos. Sci. Lett., 20, e927, https://doi.org/10.1002/asl.927.
Schulzweida, U., 2021: CDO user guide (version 2.0.0). MPI for Meteorology Doc., 215 pp., https://doi.org/10.5281/zenodo.5614769.
Singh, R., K. Tsigaridis, A. N. LeGrande, F. Ludlow, and J. G. Manning, 2022: Investigating hydroclimatic impacts of the 168–158 BCE volcanic quartet and their relevance to the Nile River basin and Egyptian history. Climate Past, 19, 249–275, https://doi.org/10.5194/cp-19-249-2023.
Skamarock, W. C., J. B. Klemp, M. G. Duda, L. D. Fowler, S.-H. Park, and T. D. Ringler, 2012: A multiscale nonhydrostatic atmospheric model using centroidal Voronoi tesselations and C-grid staggering. Mon. Wea. Rev., 140, 3090–3105, https://doi.org/10.1175/MWR-D-11-00215.1.
Son, J.-H., and K.-H. Seo, 2020: Mechanisms for the climatological characteristics and interannual variations of the Guinea Coast precipitation: Early summer West African monsoon. Atmosphere, 11, 396, https://doi.org/10.3390/atmos11040396.
Sylla, M. B., F. Giorgi, E. Coppola, and L. Mariotti, 2013: Uncertainties in daily rainfall over Africa: Assessment of gridded observation products and evaluation of a regional climate model simulation. Int. J. Climatol., 33, 1805–1817, https://doi.org/10.1002/joc.3551.
Sylla, M. B., J. S. Pal, A. Faye, K. Dimobe, and H. Kunstmann, 2018: Climate change to severely impact West African basin scale irrigation in 2°C and 1.5°C global warming scenarios. Sci. Rep., 8, 14395, https://doi.org/10.1038/s41598-018-32736-0.
Thorncroft, C. D., H. Nguyen, C. Zhang, and P. Peyrillé, 2011: Annual cycle of the West African monsoon: Regional circulations and associated water vapour transport. Quart. J. Roy. Meteor. Soc., 137, 129–147, https://doi.org/10.1002/qj.728.
Thuburn, J., T. D. Ringler, W. C. Skamarock, and J. B. Klemp, 2009: Numerical representation of geostrophic modes on arbitrarily structured C-grids. J. Comput. Phys., 228, 8321–8335, https://doi.org/10.1016/j.jcp.2009.08.006.
United Nations Convention to Combat Desertification, 2020: The Great Green Wall Initiative. United Nations, www.unccd.int/actions/great-green-wall-initiative.
Wilks, D. S., 2016: The stippling shows statistically significant grid points: How research results are routinely overstated and overinterpreted, and what to do about it. Bull. Amer. Meteor. Soc., 97, 2263–2273, https://doi.org/10.1175/BAMS-D-15-00267.1.
Xue, Y., and J. Shukla, 1993: The influence of land surface properties on Sahel climate. Part I: Desertification. J. Climate, 6, 2232–2245, https://doi.org/10.1175/1520-0442(1993)006<2232:TIOLSP>2.0.CO;2.
Xue, Y., and M. J. Fennessy, 2002: Under what conditions does land cover change impact regional climate? Global desertification: Do humans cause deserts? UCLA Rep. 4, 59–74, https://escholarship.org/uc/item/8v7011rp.
Yu, Y., M. Notaro, F. Wang, J. Mao, X. Shi, and Y. Wei, 2017: Observed positive vegetation-rainfall feedbacks in the Sahel dominated by a moisture recycling mechanism. Nat. Commun., 8, 1873, https://doi.org/10.1038/s41467-017-02021-1.
Zender, C., 2022: NCO user guide. University of California, Irvine, Doc., 453 pp., https://nco.sourceforge.net/nco.pdf.