To investigate the response of both RCA4 and HCLIM-ALADIN to horizontal resolution, we conduct a set of sensitivity experiments driven by the ERA-Interim reanalysis (denoted as ERAINT hereafter; Dee et al., 2011) at four different resolutions. These resolutions are 1.76, 0.88, 0.44, and 0.22∘ for RCA4 with the rotated coordinate system and 200, 100, 50, and 25 km for HCLIM-ALADIN with the Lambert Conformal projection. The 0.44∘ or 50 km resolution is recommended by the CORDEX experiment design and used in the CORDEX-Africa ensemble. Hereafter, the resolution in kilometres is used unless otherwise specified.

There are two approaches to set up a RCM experiment with simulations at different resolutions. The first approach is to use the same full domain (including the relaxation zone) for all simulations at different resolutions. Size of the full domain is defined by the coarsest resolution in the experiment (200 km in our case). A benefit of such an experiment setup is a consistent lateral boundary forcing for all simulations, given the same full domain. However, an unnecessary large full domain for resolutions finer than 200 km (i.e. 100, 50, and 25 km) leads to larger RCM internal variability (IV) compared to simulations at the same resolutions but with a minimum size full domain. Computational demands at the finer resolutions are also higher in the case of the large full domain. The second approach is to use different (minimum) full domains for different resolutions defined only by size of the active domain (the same for all resolutions) and a necessary relaxation zone (smaller in kilometre for higher resolution). An advantage of this approach is less IV and less computational demand for high-resolution simulations, while a shortcoming is inconsistent lateral boundary forcing (different size of the full domain). We decided to use the second approach with the minimum size of the full domain (less IV and computational demand), although we note that a perfect experiment has to include both approaches, if resources allow. The setup of the simulations at the four resolutions is identical apart from the time step (adjusted to ensure numerical simulation stability) and the size of the full computational domain with the relaxation zone (see Table 1). The relaxation zone has eight grid points in all directions and increases (in kilometres) at coarser resolution, while the interior CORDEX-Africa domain is the same.

As mentioned above, larger size of the computational domain at coarser resolution in our experiment setup may have a potential impact on the simulated climatology, leading to larger IV developed by the RCMs and weaker constraints on the ERAINT forcing. As a simple test for domain-dependent RCM IV we perform an additional experiment with RCA4 at 0.88∘ resolution taking the full computational domain from the 1.76∘ RCA4 simulation. Indeed, for the 1981–2010 climatology, seasonal mean precipitation differences between the two experiments can reach up to 1.25 mm d-1 (up to 25 %) at a few individual grid boxes, often at the edges of the tropical rain belt, although in general they stay below 0.5 mm d-1 (not shown). Seasonal mean temperature also differs with up to 1.25 ∘C regionally (not shown). We do not focus on this single additional sensitivity experiment in the study. A full set of simulations with the same full domain for all RCMs and resolutions is necessary for robust conclusions.

Another source of IV in RCMs is related to different initialization or starting time (e.g. Lucas-Picher et al., 2008; Sanchez-Gomez and Somot, 2018). We perform two additional experiments in order to see how different initialization time impacts the IV in the RCMs. Both RCA4-v1 and ALADIN at 50 km were initialized on 1 January 1980 instead of 1 January 1979 as for all other simulations in the study. It was found that the impact of the different starting time is much smaller than the impact of the larger domain. For both seasonal mean precipitation and temperature, differences between the experiments are small over the African continent, in general, less than 0.5 mm d-1 for precipitation and 0.25 ∘C for temperature (not shown). Similar to the domain-dependent sensitivity experiment above, we do not focus on these two additional initialization sensitivity experiments in the study. A full investigation of the initialization-related RCM IV needs generation of a larger (up to 10 members) ensemble for all RCMs and resolutions.

We note that in general, both regional models – RCA and HCLIM-ALADIN – were developed to operate at a range of tens of kilometres resolution, and their performance at 100 and especially at 200 km may not be optimal. A potential caveat here is that very few RCM physical parameterizations are automatically scaled to run at very coarse resolution. Thus, RCM deficiencies at the coarser resolutions may be partly related to the lack of model retuning. We think that such coarse-resolution simulations are a useful supplement to simulations at a RCM comfortable resolution zone and help us to understand RCM behaviour without additional resolution-dependent tuning. All simulations are conducted without spectral nudging similar to the CORDEX-Africa RCMs (Nikulin et al., 2012) allowing each RCM to develop its own climatology as much as possible. Analysis is done for the CORDEX-Africa domain shown in Fig. 1.

The difference between a RCM and its driving GCM can, in general, be attributed to three sources, namely (i) different resolution, (ii) different physical formulation, and (iii) artefacts of the one-way nesting approach including size of the RCM domain and application of spectral nudging (e.g. Scinocca et al., 2016). The RCA4 0.88∘ simulations and the HCLIM-ALADIN 100 km one represent a slight upscaling of ERAINT (about 0.7∘ or about 77 km at the Equator), and we refer to them as “no added value experiment” (NAVE). No resolution-dependent added value of the RCMs is expected for these NAVE simulations, and all differences between the RCMs and their driving ERAINT are attributed to different physical formulations and to the artefacts of the one-way nesting. Spectral nudging is not used in our experiment, and the one-way nesting term is basically reduced to domain configuration. In contrast, if spectral nudging is used, technical aspects of the nudging (e.g. which wavelengths should be nudged and at what altitudes) also contribute to the one-way nesting term. In practice, it is not straightforward (if possible at all) to separate the impact of different physical formulation and artefacts of the one-way nesting approach. Hereafter, we use “RCM formulation” as a term that includes both RCM physical formulation and domain-dependent RCM configuration (e.g. size of the full domain).

Observational datasets in Africa, in general, agree well for large-scale climate features but can deviate substantially at regional and local scales (Fekete et al., 2004; Gruber et al., 2000; Nikulin et al., 2012). To take into account the observational uncertainties, we utilize a number of gridded precipitation datasets. They include three gauged-based datasets: the Global Precipitation Climatology Centre, GPCC, version 7 (Schneider et al., 2014); the Climate Research Unit Time Series, CRU TS, version 3.23 (Harris et al., 2014); and University of Delaware, UDEL, version 4.01 (Legates and Willmott, 1990). All these three datasets are at 0.5∘ horizontal resolution. For the evaluation of precipitation extremes and diurnal cycle simulated by RCMs, we utilize a satellite-based precipitation dataset from the Tropical Rainfall Measuring Mission, TRMM 3B42 version 7 (Huffman et al., 2007), which is at 0.25∘ horizontal resolution and 3-hourly temporal resolution. The TRMM product starts in 1998, and for evaluation of precipitation extremes and diurnal cycle we use a shorter period (1998–2010) in contrast to 1981–2010 used for evaluation of seasonal means and annual cycle. We also note that the TRMM 3B42-v7 precipitation product provides satellite-based precipitation estimates adjusted by the GPCC gauge-based precipitation. This means that monthly mean TRMM 3B42 and GPCC precipitation are almost the same if remapped to the same resolution or averaged over a region.

ERAINT as the driving reanalysis is also used for analysis. In contrast to climate models, ERAINT precipitation is a short term forecast product, and there are several ways to derive ERAINT precipitation (e.g. different spin-up, base time, and forecast steps), which can lead to different precipitation estimates (Dee et al., 2011). ERAINT precipitation for this study is derived by the simplest method, without spin-up as in some of the previous studies (Dosio et al., 2015; Moufouma-Okia and Jones, 2015; Nikulin et al., 2012): 3-hourly precipitation uses the base times of 00:00/12:00 and forecast steps 3/6/9/12 h, while daily precipitation uses base times of 00:00/12:00 and forecast steps of 12 h. The RCMs and ERAINT represent 3-hourly mean precipitation for the 00:00–03:00, 03:00–06:00, ..., 21:00–00:00 intervals, while TRMM precipitation averages represent approximately the 22:30–01:30, 01:30–04:30, ..., 19:30–22:30 UTC intervals.

The coarsest resolution of 200 km is used as a reference resolution for spatial maps. The higher-resolution simulations are aggregated to the 200 km grid by the first-order conservative remapping method (Jones, 1999). In this way we expect that the difference among the aggregated simulations at common resolution should mainly be caused by the different treatment for fine-scale processes (Di Luca et al., 2012). For the regional analyses, such as the analysis of annual cycle, diurnal cycle, and daily precipitation intensity, we focus on four subregions, presenting different climate zones in Africa: West Africa (10∘ W–10∘ E, 7.5–15∘ N), East Africa (30–40∘ E, 15–0∘ S), southern Central Africa (10–25∘ E, 10–0∘ S), and eastern South Africa (20–36∘ E, 35–22∘ S) as defined in Fig. 1. The period 1981–2010 is used for the analysis in this study, unless otherwise specified.

In the boreal summer defined here as July–September (JAS), the tropical rain belt (TRB) associated with the intertropical convergence zone (ITCZ) is positioned to its northernmost location with the maximum precipitation north of the Equator (Fig. 2a). CRU, UDEL, and GPCC aggregated to the 200 km resolution, generally agree well with each other, with only slight local differences (Fig. 2a–c). ERAINT overestimates precipitation over Central Africa and along the Guinea coast, while underestimating it over West Africa, north of the Guinea coast (Fig. 2d). All RCA4-v1 simulations have a pronounced dry bias (Fig. 2e–h) that spatially almost coincides with the wet bias in ERAINT and increases at coarser resolution (Fig. 2e–f). RCA4-v4 shows a similar pattern compared to RCA4-v1 but substantially reduces the dry bias over Central Africa at all four resolutions (Fig. 2i–l). For both configurations of RCA4, the smallest dry bias is found at the highest 25 km resolution. At the same time, an overestimation of precipitation north of the central dry-bias region becomes more pronounced, especially for RCA4-v4. HCLIM-ALADIN, in general, shows some similarities to RCA4 with a pronounced dry bias in West and Central Africa at 200 km that is strongly reduced with increasing resolution. However, a wet bias emerges on the northern flank of the rain belt at 50 and 25 km. For JAS there is a common tendency for both RCMs to generate more precipitation at higher resolution, leading to a reduction in the dry biases over Central Africa. Such a bias reduction may be considered as a resolution-related improvement. However, the RCM simulations clearly show that the added value of higher resolution can be region-dependent. An improvement in the simulated precipitation climatology over one region corresponds to deterioration of the climatology over another region. Moufouma-Okia and Jones (2015) found a mixed response to resolution in simulated seasonal mean precipitation over West Africa. Their RCM simulations at 50 and 12 km bear a great deal of similarity with each other, while a simulation at 25 km shows wetter conditions in the Sahel and drier ones near the coastal area in the south (see their Fig. 8). In contrast, Panitz et al. (2014) found almost no difference in seasonal rainfall over West Africa between two RCM simulations at 50 and 25 km. We conclude that for both RCA4 and HCLIM-ALADIN, spatial bias patterns are similar and more related to model formulation, while magnitude of biases are more sensitive to resolution. For example, the sign of the bias pattern in our no added value RCM simulations at 100 km in JAS (Fig. 2f, j, n) is almost opposite to the sign of the bias pattern in the driving ERAINT (Fig. 2d).

In boreal winter (December–February, DJF), the TRB migrates to its most southerly position covering the latitudes from southern to Central Africa, with the maximum over southern tropical Africa and Madagascar (Fig. 3a). Similar to JAS, observational uncertainties are generally small in DJF, and there is a pronounced wet bias in ERAINT over Central Africa (Fig. 3d). At 25 and 50 km RCA4-v1 has a dipole bias pattern with an underestimation of rainfall over Central Africa and an overestimation over South Africa. At 200 km there is a pronounced deterioration in the simulated rainfall: a strong dry bias appears along the eastern coast and Madagascar, while the wet bias is amplified over large parts of southwestern Africa. At 25 and 50 km RCA4-v4 shows a large-scale dipole bias pattern similar in some degree to RCA4-v1. The RCA4-v4 biases are smaller than the RCA4-v1 ones showing an impact of the retuning (reducing mixing in the boundary layer). The behaviour of RCA4-v4 at coarser resolution is also similar to RCA4-v1. A similar strong dry bias is emerging along the eastern coast at 200 km. However, in contrast to RCA4-v1, the dry bias over the Democratic Republic of the Congo almost completely disappears at both 100 and 200 km. HCLIM-ALADIN simulates almost the same bias pattern at all resolutions, strongly underestimating rainfall over southeastern Africa and overestimating it over the Guinea coast, parts of Central Africa, and southern Africa. There is a tendency to an increase in precipitation with higher resolution in HCLIM-ALADIN: the wet biases are amplified, and the dry biases are reduced. Both RCA4 and HCLIM-ALADIN show a common feature – intensification of the dry bias along the eastern coast of Africa at 200 km. Even if both RCMs have this dry bias in common, there are also differences showing the importance of model formulation. HCLIM-ALADIN has about the same bias pattern at all four resolutions, while the RCA4 bias pattern substantially changes across the resolutions. Such resolution dependency in RCA4 may be related to the fact that RCA4 is based on a limited area model and not developed to operate at 100–200 km resolution. Contrastingly, HCLIM-ALADIN, which is based on a global model, shows more consistent results even at 100–200 km resolution. This indicates that HCLIM-ALADIN parameterizations may be better suited to work also at coarser resolution. Although, we also note that the resolution dependency of the RCA4 bias pattern over southern Africa is similar to that found for the CMIP5 GCMs (Munday and Washington, 2018). They show that the GCMs with the coarsest resolution and respectively the lowest topography have the wettest bias over the Kalahari basin and the driest bias over the southeast African coast, the Mozambique Channel, and Madagascar. Such a bias pattern is related to a smoother barrier to northeasterly moisture transport from the Indian Ocean that penetrates across the high topography of Tanzania and Malawi into subtropical southern Africa. However, in our analysis, HCLIM-ALADIN does not show such resolution-related dependency. In general, similar to JAS, the added value of higher resolution in DJF is region dependent, with higher-resolution biases reduced over one region but amplified over another.

The annual cycle of precipitation over the four subregions is shown in Fig. 4. The observed annual cycle of precipitation over West Africa depicts the West African monsoon (WAM) rainfall, with maximum precipitation in August (Fig. 4a). All observational datasets (CRU and UDEL are not shown) and ERAINT agree well with each other, with only a small underestimation of rainfall by ERAINT in June–August. In contrast to the observations, RCA4-v1 has a bimodal annual cycle with a too early onset of the rainy season (Fig. 4b). The simulated rainfall is overestimated in March–May, underestimated in July–August during the active WAM period and is well in line with the observations during the cessation of the WAM rainfall in September–November. RCA4-v4 shows a similar behaviour, but the first rainfall peak in May is reduced, and the annual cycle has a more unimodal shape (Fig. 4c). HCLIM-ALADIN, in general, shows similar features as both configurations of RCA4, although it has more similarities with RCA4-v4 (Fig. 4d). The too early onset of the rainy season is a common problem for many RCMs reported by Nikulin et al. (2012). Our results show that this is not dependent on resolution but instead related to model formulation. Higher resolution reduces the wet bias during the onset of the rainy season for RCA-v1, has no impact for RCA-v4, and amplifies the wet bias in HCLIM-ALADIN. Nevertheless, the impact of higher resolution is more consistent during the rainy season. Increasing resolution tends to increase monsoon rainfall for both RCMs, resulting in smaller dry biases and a pattern closer to the unimodal one in the observations. East and Central Africa have a bimodal annual cycle of rainfall with two peaks around November and May (Fig. 4e, i). GPCC, CRU, and UDEL (both not shown) agree well on the phase and magnitude of the annual cycle for both subregions. ERAINT has a weaker bimodality overestimating precipitation in December–February over East Africa and all year round over Central Africa with the largest wet bias during October–April. Both configurations of RCA4 fail to reproduce the bimodal annual cycle in East Africa at 200 km, underestimating precipitation all year round and showing a single rainfall peak in December (Fig. 4j, k). Increasing resolution reduces the dry bias and leads to an improvement in the shape of the annual cycle. The bimodal shape begins to appear at 100 km and becomes much closer to the observation at 50 and 25 km. Despite some mixed dry and wet biases in different seasons, the 25 and 50 km RCA4 simulations show the best agreement with the observations. In contrast to RCA4, HCLIM-ALADIN simulates the unimodal annual cycle at all four resolutions, and some signs of bimodality only appear at 25 km (Fig. 4h). Similar to RCA4, increasing resolution leads to an increase in precipitation in HCLIM-ALADIN, although a dry bias is a prominent feature from November to May in all HCLIM-ALADIN simulations. For Central Africa, the bimodality of the annual cycle is reproduced well by both RCMs at all resolutions (Fig. 4j–l). HCLIM-ALADIN maintains similar behaviour to that in East Africa, although the difference in precipitation across the resolutions is small (Fig. 4l). On the other hand, for both configurations of RCA4 in Central Africa, increasing resolution leads to decreasing precipitation during the rainy seasons, especially in January. Both RCMs strongly reduce the ERAINT wet bias even in the NAVE at 100 km. Such improvement indicates that model formulation plays a more important role than resolution over Central Africa. For eastern South Africa, the annual cycle of precipitation is unimodal with its maximum during austral summer (Fig. 4m). Similar to West Africa, uncertainties between observational datasets and reanalysis are small. RCA4 in general overestimates rainfall during the rainy season with the largest wet bias at 200 km. Surprisingly, the simulated rainfall is almost the same at 25 and 100 km, while the smallest bias is found at 50 km for both RCA4 configurations. HCLIM-ALADIN also overestimates precipitation during the rainy season at all four resolutions (Fig. 4p). However, the smallest wet bias in the HCLIM-ALADIN simulations is found at 50 and 100 km.

The diurnal cycle is a prominent feature of forced atmospheric variability with a strong impact on regional- and local-scale thermal and hydrological regimes. The diurnal cycle of precipitation in the tropics is well documented and includes a late afternoon/evening maximum over land (Dai et al., 2007). However, it is still a common challenge for GCMs (Dai, 2006; e.g. Dai and Trenberth, 2004; Dirmeyer et al., 2012), RCMs (e.g. Da Rocha et al., 2009; Jeong et al., 2011; Nikulin et al., 2012), and reanalyses (Nikulin et al., 2012) to accurately represent the diurnal cycle of precipitation.

The TRMM diurnal cycle of precipitation generally shows an increase in rainfall starting around noon with maximum reached at around 18:00 local solar time (LST) (Fig. 5). The ERAINT diurnal cycle is completely out of phase over all subregions with the occurrence of maximum precipitation intensity around local noon. A common feature of ERAINT is an overestimation of precipitation around local noon and an underestimation during the rest of the day. HCLIM-ALADIN shows exactly the same behaviour as ERAINT. Both configurations of RCA4 simulate the diurnal cycle of precipitation more accurately compared to ERAINT and HCLIM-ALADIN. The phase of the diurnal cycle, in general, is pretty well captured over all four subregions. In terms of precipitation intensity RCA4 underestimates rainfall from afternoon to morning over West (Fig. 5b, c) and Central Africa (Fig. 5j, k). Reducing mixing in the boundary layer results in flattening of the diurnal cycle over West Africa (Fig. 5b, c), while there are almost no changes over Central Africa (Fig. 5j, k). RCA4-v1 simulates very well the diurnal cycle over East Africa with only some underestimation in early morning and afternoon (Fig. 5f). RCA4-v4 improves rainfall intensity in early morning but at the same time shows a slightly larger underestimation in afternoon than RCA4-v1 (Fig. 5g). Over southern Africa the RCA4 simulations at 200 km are the closest to the observation (Fig. 5n, o), while the simulations at higher resolutions underestimate the amplitude of the diurnal cycle in the afternoon.

Figure 5 clearly shows that the phase of the diurnal cycle of precipitation in Africa does not depend on resolution but instead depends on model formulation. Both ERAINT, with the Tiedtke convection scheme (Tiedtke, 1989), and HCLIM-ALADIN, with the Bougeault scheme (Bougeault, 1985), trigger precipitation too early during the diurnal cycle, while both configurations of RCA4 with the same Kain–Fritsch (KF) scheme (Bechtold et al., 2001) simulate a much more realistic diurnal cycle. It has previously been shown that the KF scheme is able to reproduce late afternoon rainfall peaks for the regions where moist convection is governed by the local forcing, for example in the southeast US (Liang, 2004) and in tropical South America and Africa (e.g. Bechtold et al., 2004; Da Rocha et al., 2009). Nikulin et al. (2012) also found that a subset of RCMs that employ the KF scheme show an improved representation of the phase of the diurnal cycle in Africa. Our results indicate that the impact of resolution is only seen in the amplitude of the diurnal cycle. However, such impact is not homogeneous across the subregions and the RCMs. For HCLIM-ALADIN, increasing resolution leads to increasing rainfall intensity in all regions but eastern South Africa. RCA4 shows a similar behaviour over West Africa, while there is a mixed response over East and Central Africa. These findings are in line with previous studies investigating resolution effects for GCMs (Covey et al., 2016; Dirmeyer et al., 2012) and for RCMs (Walther et al., 2013). In coarser-scale models (e.g >10 km), increasing resolution only leads to changes in the magnitude but not in the phase of the diurnal cycle of precipitation over land.

Nevertheless, studies conducting sensitivity experiments using resolutions finer than 10 km do find improvements in the representation of the phase (Dirmeyer et al., 2012; Sato et al., 2009; Walther et al., 2013).

Figure 6 shows the empirical probability density function (PDF) of daily precipitation intensities over the four subregions. The TRMM7-0.25 dataset, aggregated to the common 1.76∘ resolution (TRMM7-1.76), as expected has a shorter right tail with no precipitation intensities larger than 100 mm d-1 and higher frequency for lower intensities less than 25 mm d-1 (Fig. 6a, e, i, m). The two TRMM7 PDFs provide reference bounds for datasets with resolution between 0.25 and 1.76∘. However, uncertainties in gridded daily precipitation products in Africa are large (Sylla et al., 2013a), and we take the TRMM bounds as an observational approximation focusing more on differences in the simulated PDFs across the four resolutions. Over West, East, and Central Africa ERAINT overestimates the frequency of low (<10 mm d-1) and extremely high (>150 mm d-1) intensities, while it underestimates the frequency of precipitation intensities in between (Fig. 6a, e, i), especially over West Africa (Fig. 6a). In South Africa (Fig. 6m) ERAINT represents the frequency of daily mean precipitation more accurately compared to the other three regions but shows almost no events with more than 150 mm d-1 in contrast to the observations. Both RCMs, in general, have the same tendency to generate more higher-intensity precipitation events with increasing resolution over all four subregions. In West Africa RCA4-v1 strongly underestimates the frequency of intensities with more than 20 mm d-1 at 200, 100, and 50 km (Fig. 6b). A substantial improvement appears only at 25 km where the right tail of the PDF extends up to 250 mm d-1, although the frequency of precipitation events from about 50 to 150 mm d-1 is still underestimated.

The RCA4-v4 configuration markedly reduces the RCA4-v1 biases and shows more realistic PDFs at all four resolutions (Fig. 6c). The RCA4-v4 50 km simulation generates precipitation events up to 250 mm d-1, strongly contrasting the RCA4-v1 simulation at the same resolution (no events for more than 100 mm d-1). However, RCA4-v4 overestimates frequencies of high intensities at 25 km. Such sharp difference between two configurations of RCA4 at the same resolution shows that model formulation also plays an important role for accurately reproducing daily precipitation. Over West Africa all HCLIM-ALADIN simulations overestimate the frequency of low precipitation intensities (less than 10 mm d-1) and underestimate the frequency of intensities in the range of 10–150 mm d-1 (Fig. 6d). Similar to RCA4, higher resolution leads to more high-intensity precipitation events in the HCLIM-ALADIN simulations.

However, RCA4 and HCLIM-ALADIN behave in a different way with increasing resolution. Both RCMs change the PDFs by adding more higher-intensity precipitation events, extending the righthand tail towards higher intensities. In addition, RCA4 also increases the frequency of medium- and high-intensity events, especially going from 50 to 25 km. In East Africa both RCA4 configurations reproduce the observed PDFs almost perfectly (Fig. 6f, g). All four resolutions are located within the TRMM-1.76 and TRMM-0.25 boundaries, and the coarsest and finest resolutions coincide with the respective TRMM PDFs. Contrastingly, HCLIM-ALADIN strongly underestimates the frequency of precipitation events with more than 20 mm d-1 (Fig. 6h) over East Africa, and even the highest 25 km resolution is located below the coarse TRMM-1.76 dataset. In Central Africa both RCMs overestimate the occurrence of intensities less than 20 mm d-1 (Fig. 6j, k, l), especially HCLIM-ALADIN (Fig. 6l), and strongly underestimate the frequency of higher-intensity events. The PDFs at all four resolutions for both RCMs are located below the coarsest TRMM-1.76 PDF. We note that observational uncertainties in precipitation are very large over Central Africa, and we should be careful in the interpretation of Fig. 6j–l. Seasonal mean precipitation, for example, can differ by more than 50 % across different observational datasets (Washington et al., 2013). Additionally, the TRMM dataset is scaled by the gauge-based GPCC precipitation product, while almost no long-term gauges are available in the region (Nikulin et al., 2012). In eastern South Africa RCA4 and HCLIM-ALADIN simulate the precipitation PDFs quite accurately (Fig. 6n–p). An interesting detail is that the 50 km HCLIM-ALADIN simulation shows higher frequency for intensities in the range of 50 to about 200 mm d-1 than the 25 km simulation.

In general, we see the improvement of simulated daily rainfall intensities with increasing resolution across the African continent. There are many studies showing a similar resolution-dependent improvement over both complex terrains and flat regions (e.g. Chan et al., 2013; Huang et al., 2016; Lindstedt et al., 2015; Olsson et al., 2015; Prein et al., 2016; Torma et al., 2015; Walther et al., 2013). Our results are in agreement with the above studies and confirm increasing fidelity of simulated daily rainfall intensities with increasing resolution.