1. Introduction
Urban areas can influence local precipitation due to the combined effects of the urban heat island (UHI), surface roughness, and aerosols (Oke et al. 2017). Urban precipitation, especially when intense, induces flooding, which in turn negatively impacts human livelihoods, economies, and ecosystems (IPCC 2023). In the last few decades, extensive studies have been conducted to describe how urban areas interact with local precipitation and underlying physics. Horton (1921), a pioneer in this field, observed the increase of thunderstorms in New York City (New York) and Rhode Island and attributed this change to the urban effect (Horton 1921). Later, a famous large-scale observation campaign, the Metropolitan Meteorological Experiment, showed that UHI increased the frequency and amount of summer precipitation in Saint Louis, Missouri, and its downwind area (Huff and Changnon 1972; Changnon 1981). This finding has been supported by subsequent observational studies worldwide using both in situ measurements and state-of-the-art satellite observations (Yonetani 1982; Westcott 1995; Orville et al. 2001; Grady Dixon and Mote 2003; Burian and Shepherd 2005; Mote et al. 2007; Niyogi et al. 2011; Wang et al. 2021).
Advances in atmospheric and urban numerical modeling allow researchers to investigate the precipitation sensitivity to urban effects and shed light on the detailed physical processes (Thielen et al. 2000; Rozoff et al. 2003; Burian and Shepherd 2005; Gero and Pitman 2006; Baik et al. 2007; Shem and Shepherd 2009; Lin et al. 2011; Kusaka et al. 2014, 2019; Doan et al. 2021; Simón-Moral et al. 2021). Numerical modeling has primarily advanced urban-precipitation studies over areas without fine-scale observation systems to assess urban–rural variability. Hence, robust and detailed knowledge of the physical processes involved in urban precipitation has been produced. The dominant theory is that an urban area enhances the convergence of low-level winds due to UHI, invigorating thermals (Rozoff et al. 2003; Baik et al. 2007; Shem and Shepherd 2009) that result in convection initiation and precipitation intensification over a city and its downwind region. Zhang et al. (2019) added that urban roughness might contribute to prolonged rainfall. Interaction between urban effects and other mesoscale mechanisms has also been investigated (Gero and Pitman 2006; Lin et al. 2011; Kusaka et al. 2014, 2019; Argüeso et al. 2016; Freitag et al. 2018; Doan et al. 2021). For example, enhanced pressure gradients and frictional convergence in urban areas may pull sea breezes more strongly into coastal cities, transporting more water vapor onshore, thus increasing the potential for convection and precipitation (Argüeso et al. 2016; Doan et al. 2021).
Despite its decades-long history, there remain many challenges in urban-precipitation research (Liu and Niyogi 2019; Qian et al. 2022). First, despite commonly held assumptions, there is no consensus on the urban effect on precipitation. There are reports that claimed urban areas are responsible for a reduction in precipitation. Kaufmann et al. (2007) found that urbanization reduced local precipitation in the Pearl River Delta, China (Kaufmann et al. 2007). By analyzing long-term rainfall data, Zhang et al. (2009) revealed that the rapid urban expansion in Beijing since 1981 is statistically correlated to summer rainfall reduction in the city’s northeast areas (Zhang et al. 2009). They argued that urban expansion reduces evaporation, increases surface temperatures and sensible heating, then deepens the boundary layer. This leads to less water vapor and more mixing of water vapor in the boundary layer, and hence less convective available potential energy and more convective inhibition energy. Hamdi et al. (2012) pointed out that incorporating urban effect into a numerical model might induce both decrease and increase in precipitation, which are also seasonally and spatially dependent (Hamdi et al. 2012). A recent extensive review conducted by Liu and Niyogi (2019) pointed out another problem in urban precipitation studies that is related to geographical bias of target areas. Most existing studies have focused on specific urban regions, especially in midlatitude countries like the United States and China. The lack of geographical diversity in urban precipitation analysis may limit a comprehensive understanding of the issue. According to the authors’ knowledge, no study has addressed the urban-precipitation effect over an African megacity, despite recent efforts addressing the UHI effect in the region (Ojeh et al. 2016; Bassett et al. 2020).
More challenges lie in understanding the relationship between urban effects and extreme precipitation events (EPEs, defined as heavy precipitation events for a given duration). Because EPEs rarely occur, it is difficult to find sufficient samples for a reliable analysis. Also, EPEs are often associated with synoptic atmospheric circulation; thus, tracking urban signals is exceptionally difficult. This explains why extreme urban precipitation has been less studied until recently, thanks to progress in data collection and numerical modeling. The evident urban-enhancement effect of EPEs is detected from historical observation data. For example, Golroudbary et al. (2017) show that urban areas may have affected the extreme precipitation patterns across the Netherlands (Golroudbary et al. 2017). They found the EPEs (in terms of monthly maxima) of observed (1961–90) daily precipitation was greater in the urban areas than in the rural areas. Li et al. (2019) arrived at the same conclusion for the urban effect by analyzing observational data (1960–2010) in Shanghai and Guangzhou, China. From a different perspective, Pathirana et al. (2014) conducted numerical simulations and showed that the intensity of EPEs increases significantly with the expansion of urban areas in Colombo (Sri Lanka), Dhaka (Bangladesh), and Mumbai (India). Li et al. (2020) took the same approach and found that urbanization increased the 95th percentile precipitation value by 30% in Kuala Lumpur, Malaysia. Luong et al. (2020) tracked the urban signal from 10 heavy rainfall events in Jeddah, Saudi Arabia, using a numerical approach and found that the presence of cities increased precipitation by an average of 26%. Nonetheless, the urban-weakening effect of EPEs is also reported. Zhang et al. (Zhang et al. 2019) statistically analyzed the observed data in Beijing, China, and concluded that urbanization reduced the magnitude and frequency of EPEs but extended the continuous rainfall days.
Together with an increasing trend in studying urban EPEs, there is also a rising number of research studies targeting cities in the tropical region. Tropical areas are characterized by the hot, humid atmosphere and the dominance of convective activities, which occur at a relatively small scale corresponding to the size of a large city. Studies are specifically seen in South Asia, such as Kolkata (India; Mitra et al. 2012) and Mumbai (Shastri et al. 2015), and in Southeast Asia, such as Kuala Lumpur (Ooi et al. 2017; Li et al. 2020), Jakarta, Indonesia (Argüeso et al. 2016), and Singapore (Doan et al. 2021; Simón-Moral et al. 2021). Doan et al. (2021) showed that precipitation increases in urban areas, especially during the intermonsoon season where the relatively weak prevailing wind allows the dominance of convective processes. A strong consensus among studies is that the urban effect in the tropics is likely stronger than that in midlatitude areas. The impact of urban areas could reach 20%–30% (Argüeso et al. 2016; Doan et al. 2021). These studies argued that as convection is the dominant rainfall process, urban areas in the tropics have a stronger effect than counterpart cities in midlatitudes where rainfall comes from many sources, for example, frontal or low pressure (Oke et al. 2017).
This study aims to address urban extreme precipitation in a tropical coastal megacity. We take Lagos, Nigeria, as a case study. Lagos is one of the most populated cities in the African continent, showing an extraordinarily high rate of urbanization, and it is predicted to be the world’s largest city by 2100 with more than 88 million inhabitants (Hoornweg and Pope 2017). Located close to the southern coastline of Nigeria, the city is exposed to the interaction between the African monsoon climate and the local land–sea-breeze impact resulting extreme rainfall events in the area. Rapid urbanization, accompanied with well-known urban heat island effect, increases the complexity of the local extreme rainfall processes. Also, inadequately planned urban expansion (Faisal Koko et al. 2021) implies the aggravated vulnerability of infrastructure and livelihoods associated with extreme rainfall events. Focusing on Lagos, we address two questions: Does the urban area modify the extreme precipitation occurring over the city? What physical processes are responsible for these changes, if any? To do so, we employ the state-of-the-art numerical modeling system, the Weather Research and Forecasting (WRF) Model, within an ensemble simulation framework with multiple combinations of different physical schemes and boundary/initial conditions to avoid the intrinsic uncertainty/noise in modeling for detecting robust urban signals. The ensemble simulation framework allows us to identify which model configuration works best for urban-precipitation modeling in the region; such information could benefit subsequent studies addressing the same issue. Besides the scientific contribution, this study has broader implications for studying urban hazards in African megacities. This endeavor is a critical step toward achieving the U.N. Sustainable Development Goals (United Nations 2015), especially, number 11 “make cities and human settlements inclusive, safe, resilient and sustainable,” and number 13 “take urgent action to combat climate change and its impacts” based on understanding the patterns and physical mechanisms of urban extreme precipitation to support hazard mitigation through improving urban sewerage system and urban flooding management system.
2. Method
a. Study area
We select Lagos, the capital city of Nigeria, as our case study. Lagos is one of the largest cities in Africa. Its population was about 5.2 million in 1991 and grew to about 13 million by 2018 (United Nations 2018). Lagos is predicted to be the world’s largest city, with about 88.3 million inhabitants, by 2100 (Hoornweg and Pope 2017). The city is characterized as a tropical savanna climate, according to the Köppen climate classification (Peel et al. 2007), with a rainy season from April to October and a dry season from November to March (Bassett et al. 2020; also see Fig. 1). The city is strongly influenced by the West African monsoon, which is characterized by the African easterly jet in the midtroposphere and the African easterly waves (Hourdin et al. 2010; Sylla et al. 2013). The African easterly waves are essential drivers of convection and precipitation patterns over the region (Diedhiou et al. 1998). Lagos’s terrain is mainly flat and is bordered by the Gulf of Guinea to the south and the Lagos lagoon to the east.
Study area and background climate: (a) Map of Lagos and surrounding area. (b) Climograph of Lagos from in situ observations at Murtala Muhammed International Airport (6.58°N, 3.32°E). Data were accessed from NCEI’s Climate Data Online (https://www.ncei.noaa.gov/cdo-web/). (c) Distribution of land use/land cover around the Lagos region inside the innermost domain. The line connecting P1 and P2 shows the cross-sectional area along which subsequent sea-breeze analysis is conducted.
Citation: Journal of Applied Meteorology and Climatology 62, 2; 10.1175/JAMC-D-22-0048.1
b. Detection of extreme precipitation events and observed data
An extreme precipitation event for which ensemble simulations are performed is detected using a satellite-based product and in situ observation data. The satellite-based precipitation product, the Climate Prediction Center (CPC) morphing technique (CMORPH), consists of global, high-resolution satellite precipitation estimates (Joyce et al. 2004). The data have been bias corrected and reprocessed. First, the purely satellite-based global precipitation data are integrated from all available passive microwave measurements aboard low-Earth orbiting platforms. Bias in these integrated precipitation data is then removed through comparison with CPC daily gauge analysis over land and the Global Precipitation Climatology Project (GPCP) analyses over ocean (Xie et al. 2019). Data are reprocessed on a global grid spacing of 8 km × 8 km with the temporal interval of 30 min from January 1998 until present. The CMORPH data are used for studying rainfall climate in cities (Wang et al. 2021) or for numerical modeling evaluation (Zhong and Yang 2015). In situ data are daily rain gauge data collected from the international airport weather station Murtala Muhammed (6.58°N, 3.32°E). We evaluate the quality of CMORPH by comparing it with in situ observations. The results show that CMORPH overestimates by 19% the annual average rainfall over Murtala Muhammed International Airport (Table 1). For wet months, that is, April, May, and June, the overestimation is 40%, 5%, and 33%, respectively. The lack of subdaily in situ data prevents us from evaluating the diurnal variation of CMORPH.
Comparison of rainfall data between in situ observation (OBS) at Murtala Muhammed International Airport and satellite-driven CMORPH product from January 2001 to December 2019 for April, May, and June.
Next, we detect the extreme rainfall events (on a daily basis) from CMORPH and in situ datasets. A 19-yr period (1 January 2001–31 December 2019) is examined. First, we detect the 100 wettest days from two datasets; then, we keep overlapping days of the two sets. Note that a 1-day window is implemented to avoid miscounting events with a time lag in either dataset. Then, 22 overlapping extreme rainfall events remain (Table 2). Eight events are recognized in wet months (April–June). The event on 30 May 2006, is selected as a case study after several test simulations with WRF (explained later). This day, rainfall began around Lagos in the afternoon; at 1300 UTC (1400 local time), it exceeded 30 mm h−1 over Lagos (see Fig. 2).
Description of the rainfall event on 30 May 2006: (a) Spatial distribution of CMORPH rainfall and its evaluation in time. (b) Time series of Lagos [black-outlined box in (a)] averaged rainfall.
Citation: Journal of Applied Meteorology and Climatology 62, 2; 10.1175/JAMC-D-22-0048.1
Detected extreme rain events.
c. Numerical model configuration and ensemble experiment design
1) Weather Research and Forecasting and urban models
We use the WRF Model, version 3.9.1 (Skamarock et al. 2008), to simulate the rainfall during the extreme event of interest. To assess and reduce simulation uncertainty, we employ an ensemble approach of multiple simulations with different physical schemes and boundary/initial conditions. The WRF Model is a fully compressible nonhydrostatic model with various physical options. WRF has been applied for diverse purposes, ranging from numerical forecasting to researching local climate and future projections. We configure three nested simulation domains, with grid spacing of 27, 9, and 3 km, respectively, centered at Lagos (6.45°N, 3.39°E) (Fig. 1 and Table 3). We conduct two simulation groups, one with the urban area (identified by MODIS land type data) and the single-layer urban canopy model (UCM) activated, herein called control (CTRL) simulations, and another with no urban area (replaced by forest land cover, which is dominant non-built-up land use over the region) and no urban physics (NoURB).
Model configuration, and scheme setting.
2) Ensemble setting
Simulation ensemble is conducted, in which different physical schemes and boundary/initial conditions are used. Employing an ensemble serves two purposes: one is to reduce uncertainty in simulated rainfall results; the second is to investigate which configuration works optimally for the region, given the lack of information on urban precipitation modeling studies in Nigeria and Africa generally. For each experiment, that is, CTRL or NoURB, a total of 16 ensemble simulations are conducted (Table 4). Such 16 simulations are the result of the combination of two boundary/initial conditions [the Final operational global analysis dataset produced by the National Centers for Environmental Prediction (NCEP) at the National Ocean and Atmospheric Administration (NOAA) (NCEP-FNL) and the fifth major global reanalysis produced by the European Centre for Medium-Range Weather Forecasts (ECMWF) (ERA5)].
Ensemble members. Sixteen ensemble members are the combinations of two initial and boundary conditions (IBC), two microphysics schemes, two PBL schemes, and two initial times. Note that the ensemble was implemented for both the control and nonurban simulations; thus, a total of 32 simulations were conducted.
One of the two boundary/initial conditions is NCEP-FNL (https://rda.ucar.edu/datasets/ds083.2). NCEP-FNL data have a horizontal resolution of 1° (about 100 km) and a temporal resolution of 6 h. On the other hand, ERA5 data have a spatial resolution of 0.25° (approximately 30 km) and a temporal resolution of 1 h (Dee et al. 2011). Two cloud microphysics schemes were used: WRF double-moment 6-class scheme (WDM6; Lim and Hong 2010) and WRF single-moment 6-class scheme (WSM6; Hong and Lim 2006). These schemes are the top two that Song and Sohn (2018) found to reproduce observation data from TRMM satellites with WRF (Song and Sohn 2018). Two PBL schemes include a nonlocal scheme, Yonsei University scheme (YSU; Hong et al. 2006), and a local and nonlocal combined scheme, Asymmetric Convection Model 2 (ACM2; Pleim 2007). Two start times include 6 and 12 h before the onset of the rainfall events are used. Jee and Kim (2017) concluded that a lead time of 6 h or less is optimal for reproducing local torrential rainfall events around Seoul, South Korea (Jee and Kim 2017), whereas Luong et al. (2020) simulated 10 extreme precipitation events over Jeddah with a 12-h model spinup period and found that the WRF Model reproduced the observations satisfactorily (Luong et al. 2020).
Other physical schemes are kept the same for all simulations. Those schemes include the radiation scheme RRTMG (Mlawer et al. 1997) and the land surface scheme Noah LSM coupled with a single-layer urban-canopy scheme (Kusaka et al. 2001; Kusaka and Kimura 2004; Chen et al. 2011). The urban-canopy scheme allows the model to consider the heat and momentum flux exchange between the urban surface (e.g., building wall, roof, or road) with the air, thus improving the representation of urban effect to the atmosphere. However, the urban scheme also requires the local urban morphology information and anthropogenic heat release (the estimation method will be addressed in the following section). For the cumulus scheme, we applied the Tiedtke scheme (Tiedtke 1989) to the first and second domains (with grid spacing of 27 km × 27 km, and 9 km × 9 km, respectively). We do not use the cumulus scheme for the innermost domain 03 (with grid spacing of 3 km, which is at convection-permitting scale).
3) Estimation of urban parameters
Estimation of urban parameters used in urban canopy scheme.
3. Results and discussion
a. Tracing urban footprint
The simulated result is compared with the satellite product CMORPH to assess the model’s performance. The initiation and evolution of the rain (in and surrounding Lagos) from simulations and CMORPH are tracked and compared with each other. Members whose results match well with CMORPH are selected for further analysis. Figure 3 illustrates the results for 3-h accumulated precipitation at its peak (i.e., 1500–1700 UTC accumulated precipitation) from four best-performing members (Nos. 3, 4, 7, and 14). The selection of the four is based on both subjective evaluation with comparing the evolution of the rainfall event in time and space with that from CMORPH, and objective evaluation using the Pearson correlation coefficient between the two datasets. As shown in Fig. 3, the best four members reproduce the spatial distribution and magnitude of the rainfall event, which is approximately 40–60 mm (3 h)−1 over the area of Lagos, in reference to the CMORPH data. Among the four members, members 3, 4, and 7 were driven by the NCEP-FNL data, whereas member 14 was driven by ERA5. (The results for all members can be seen in Fig. A1 of the appendix.) Our results suggest that the model is more sensitive to boundary/initial conditions than changing physical schemes. Simulations with NCEP-FNL perform better in terms of rainfall amount, whereas those driven by ERA5 tend to have dry bias; the rain is more local and likely caused by sea-breeze impact. To preserve the significance and validity of the following analysis, we use only the best four ensemble members (numbers 3, 4, 7, and 14) for further analysis. Note that the best four have different physics options for PBL and cloud microphysics, which leads to a reasonable model ensemble that captures the variability of physical schemes for key processes. We also assess the results averaged for all 16 members. The pattern of urban effect on rainfall, and relevant physical processes, which are seen in 4-member analyses, are preserved with 16-member analyses but to a lesser extent because of the abovementioned dry bias existing in some members (see Fig. A2 of the appendix).
Results verification: (a) Spatial distribution of peak rainfall (accumulated for 1500–1700 UTC) from the best four members (Nos. 3, 4, 7, and 14) (the full version with all members is shown in Fig. A1 in the appendix). (b) As in (a), but for the CMORPH result.
Citation: Journal of Applied Meteorology and Climatology 62, 2; 10.1175/JAMC-D-22-0048.1
Our primary purpose is to learn how the urban area modifies (increases/decreases) the extreme rainfall over and around Lagos. In the following analysis, we compare and show the ensemble mean results between the best four CTRL simulations and the corresponding four NoURB sensitivity simulations (using the same boundary/initial conditions and physics schemes but without urban representation as the best four CTRL simulations). Figure 4 illustrates the spatial distribution of rainfall (at its peak) from both CTRL and NoURB and the difference between them, and Fig. 5 shows the corresponding urban effect from the Hovmöller diagram perspective of rainfall evaluation in time and space. Overall, the modeled results show the significant modification of rain around the city due to the urban effect. The urban footprint is noticeably detectable as the urban area shows ∼30% higher peak rainfall intensity in CTRL than NoURB. Quantitatively, the averaged difference between CTRL and NoURB over the urban area is noticeable (Fig. 4d). CTRL generated more than 32 mm (3 h)−1 higher than 25 mm (3 h)−1 by NoURB, implying that approximately 25%–30% increased rain rate is attributable to the urban effect. Interestingly, the results show the same timing (around 1600 UTC) of peak rainfall in both CTRL and NoURB, leading to speculation that the urban area enhances the existing convection rather than initiating a new one.
Urban footprint on spatial distribution of rainfall event at 1600 UTC: Spatial distribution of peak time (ensemble mean) from the (a) control run (CTRL) and (b) nonurban (NoURB) best four ensemble members. (c) The difference between CTRL [in (a)] and NoURB [in (b)]. (d) Time series of the spatial average over Lagos.
Citation: Journal of Applied Meteorology and Climatology 62, 2; 10.1175/JAMC-D-22-0048.1
Hovmöller diagram for development and propagation of rainfall: Results (ensemble mean) for the (a) CTRL and (b) NoURB best four ensemble members; the x axis indicates the line P1–P2 (see Fig. 1c), and the y axis indicates the time (UTC). (c) The difference between CTRL [in (a)] and NoURB [in (b)].
Citation: Journal of Applied Meteorology and Climatology 62, 2; 10.1175/JAMC-D-22-0048.1
Note that the rainfall change (between CTRL and NoURB) is characterized by both “positive” and “negative” extrema outside the city. These extrema appear to organize in forms of wave propagation. The possible explanation for these extrema is what we call a “UHI wave.” Urban thermal effect could directly reinforce the convective processes by attracting and converging moisture from surrounding area that consequently enhancing extreme rainfall events over urban areas. This leads to the loss of moisture supply and convective energy in peripheral areas, which then suffer rainfall deficit (as negative extrema seen). Also, the enhanced convection over an urban area could trigger a secondary convection cell, especially when the environment has enough moisture and convective energy to afford this. It is a reason why one can see positive extrema scattered in between negative areas (Fig. 4c). Because convective processes at kilometer-scale simulations are quasi random and uncertain, when and where the secondary convection triggers are sensitive to how physical schemes, forcing data, and nested domains are selected and set up.
b. Underlying physical mechanisms
Here, various meteorological variables are analyzed to understand why CTRL, as opposed to NoURB, generates more precipitation around the city. First, we examine the change in surface-heat budget because the primary difference between CTRL and NoURB lies in the change in land-use/land-cover type, which is essential in modulating heat balance. Figure 6 shows the spatial distribution of sensible heat flux QH an hour before the peak precipitation. The results show the extra sensible heating and the reduction of latent heating over the urban area in CTRL when compared with NoURB. The addition of QH is above 75 W m−2, corresponding to the same decreasing amount of QE (not shown). This change in sensible and latent heat is considered to be the typical UHI effect. It is commonly known that the change in land use/cover from vegetation to built-up areas reduces evapotranspiration, resulting in an increase in sensible heating, which in turn, modifies the atmospheric physical processes through land–atmosphere interactions (e.g., Doan and Kusaka 2016).
Urban impact on sensible heat flux (W m−2) at 1600 UTC: Results (ensemble mean) for the (a) CTRL and (b) NoURB best four ensemble members. (c) The difference between CTRL [in (a)] and NoURB [in (b)].
Citation: Journal of Applied Meteorology and Climatology 62, 2; 10.1175/JAMC-D-22-0048.1
The increase in upward sensible heat flux over the urban area is a reason for the enhanced UHI effect, which is reflected by enhanced surface air or skin temperature. For example, Fig. 7 shows the spatial distributions of surface air temperature, and Fig. 8 shows the PBL height (PBLH) from both CTRL and NoURB an hour before the rainfall peak. Over the urban area, CTRL has surface air temperature and PBLH higher than NoURB by 1.5 K and 150 m, respectively. The higher PBLH is a clear indicator of enhanced instability over the urban area, favoring thermals to be generated and developed, especially when the background atmosphere is already unstable; this can enhance convection and precipitation. The ground pressure around Lagos is 0.15 hPa lower in CTRL than in NoURB (not shown). Therefore, it is suggested that the city causes enhanced convective activity and enhanced convergence of the surface winds.
As in Fig. 6, but for urban surface air temperature (K).
Citation: Journal of Applied Meteorology and Climatology 62, 2; 10.1175/JAMC-D-22-0048.1
As in Fig. 6, but for PBLH (m).
Citation: Journal of Applied Meteorology and Climatology 62, 2; 10.1175/JAMC-D-22-0048.1
Previously, we discussed that the UHI effect could strengthen mixing over the urban area, thus encouraging convection and horizontal convergence. However, we need more information about low-level water moisture convergence to determine the key processes. Figure 9 shows the results for water vapor flux at 1600 UTC, with the clear convergence of air moisture over the urban area. The increased water vapor flux due to the urban effect is about 20–30 m s−1 g kg−1. The area over the south coast and nearby sea area also shows a noticeable increase of air moisture, and the direction of the enhanced water vapor flux points inland. This indicates that the UHI not only strengthens convergence over the urban area (due to extra sensible heat flux) but also boosts the land/sea temperature contrast. This contrast results in boosting sea breeze that brings additional moisture. Figure 10 further confirms this hypothesis by illustrating the obvious enhancement of sea breeze. It is also clear that the updraft region is seen to the center of Lagos restricting penetration of the sea breeze farther inland. The fact that sea breeze, which is aided by overheating over an urban area, is prevented from farther inland penetration by the UHI circulation is also shown in many previous studies (Kusaka et al. 2000, 2019; Argüeso et al. 2016; Doan et al. 2021). The moisture convergence that resulted in convection and rainfall is more likely to occur over the urban area than the downwind area. Figure 11 shows the time series of precipitation and associated atmospheric variables precipitable water (PW), convection available potential energy (CAPE), and convection inhibition (CIN), together with atmospheric soundings for a grid cell above the urban footprint of Lagos for each ensemble member from the best four. The results for PW show higher values in CTRL than those in NoURB in a few hours before peak rainfall (also see Fig. 12 for ensemble average). This supports the argument that the moisture convergence processes play a larger role than the loss of a moisture source from reduced evapotranspiration in the case of urban land cover versus forest land cover. Also, the atmospheric instability is higher in CTRL than NoURB during the peak rainfall time in sounding plots, consistent with the conclusion deduced from PBLH analysis mentioned above.
As in Fig. 6, but for surface water vapor flux (m s−1 g kg−1).
Citation: Journal of Applied Meteorology and Climatology 62, 2; 10.1175/JAMC-D-22-0048.1
Cross section of wind along the vertical plane P1–P2 (see Fig. 1c for more detail) at 1600 UTC: Results (ensemble mean) for the (a) CTRL run and (b) NoURB. Contours indicate the vertical wind W. The wind vector has two dimensions; i.e., one is the vertical and another is along the P1–P2 plane. (c) The difference between CTRL [in (a)] and NoURB [in (b)].
Citation: Journal of Applied Meteorology and Climatology 62, 2; 10.1175/JAMC-D-22-0048.1
Time series of precipitation, PW, CAPE, and CIN and the atmosphere soundings at precipitation-peak time for a grid cell above the urban footprint of Lagos for each member from the best four [Nos. (a) 3, (b) 4, (c) 7, and (d) 14] showing comparisons between CTRL and NoURB.
Citation: Journal of Applied Meteorology and Climatology 62, 2; 10.1175/JAMC-D-22-0048.1
Time series of hourly precipitation (label Prcp) and precipitable water (label PW) from a grid cell over the urban footprint of Lagos from 3 h before to 3 h after the precipitation peak. The results are averaged among the best four members.
Citation: Journal of Applied Meteorology and Climatology 62, 2; 10.1175/JAMC-D-22-0048.1
c. Discussions
It is generally accepted that the model performance on extreme precipitation is strongly dependent on model configurations, physical schemes, target regions, and forecasting times (Tewari et al. 2021). That is the challenge of this study, as the first attempt to investigate the urban extreme precipitation in Lagos. We employ the ensemble simulation approach to address the uncertainty intrinsic in precipitation modeling. We also aim to find a suitable configuration for simulating precipitation over the region. We show that the ensemble produces a wide variation of results. Specifically, the boundary and initial data selection influences the simulation accuracy more than the physical schemes do. In this study, the rain event is reproduced better with NCEP-FNL than ERA5 data. Note that the rainfall modeling is very sensitive to initial conditions if the simulation period is short at a scale of days. It is commonly understood by the numerical weather forecasting community that the initial conditions, specifically moisture amount and instability, strongly impact the final outcomes.
Physical scheme selection also influences the model’s performance, but to a lesser extent. We show that cloud microphysics does not influence much; however, the PBL schemes do. ACM2 performed better than YSU in our case study. This result differs from studies that concluded that nonlocal models are optimal (Shin and Hong 2011) but is consistent with Gunwani and Mohan (2017), who demonstrated the high climate versatility of ACM2. However, it is important to keep in mind that our results are limited to a single case study. More simulations are needed to generalize the conclusion for Lagos, or for other African cities. The time of model initialization is also important when reproducing the rainfall event. We demonstrate good reproducibility for urban extreme precipitation in simulation initialized at 6 h before the event onset for NCEP-FNL, and 12 h before the event onset for ERA5. This suggests that less spinup time is needed to generate rainfall if we have enough “wet” initial conditions and vice versa.
For cumulus schemes, we use the New Tiedtke scheme for model domain 01 and 02. There remains the question about its impact on simulated outcomes. To answer this, we conducted additional simulations with Kain–Fritsch (for domains 02 and 03 only) for the best four ensemble members of CTRL and NoURB (thus, total eight additional runs were conducted). The results show that, even though using different cumulus schemes causes large difference between total accumulated rainfall amount across all domains, the urban impact on the rainfall event is well preserved and scaled, especially with the same relative proportion of the urban impact (see Figs. 13 and 14). This result demonstrates the robustness of the conclusion of this study regardless of cumulus schemes used.
Difference in absolute values of accumulated rainfall [spatially averaged over domains (a) 01, (b) 02, and (c) 03] with using different cumulus parameterization schemes (i.e., New Tiedtke and Kain–Fritsch).
Citation: Journal of Applied Meteorology and Climatology 62, 2; 10.1175/JAMC-D-22-0048.1
Indifference in urban footprint with using difference cumulus parameterization schemes: Results from simulation using the (top) New Tiedtke and (bottom) Kain–Fritsch schemes showing spatial distribution of rainfall results (accumulated precipitation for 1500–1700 UTC) from (left) CTRL (4-member ensemble mean) and (left center) NoURB and (right center) the difference between the two. (right) The corresponding time series of spatial averaged values for Lagos.
Citation: Journal of Applied Meteorology and Climatology 62, 2; 10.1175/JAMC-D-22-0048.1
Our study shows that 20%–40% of the peak precipitation over the city is attributed to the urban effect. This finding is consistent with those of (sub)tropical cities (Wang et al. 2015; Argüeso et al. 2016; Li et al. 2020; Luong et al. 2020; Doan et al. 2021). Our study solidifies the argument that the urban effect on rainfall in the tropics is stronger than that in midlatitude cities. For example, in their extensive review, Liu and Niyogi (2019) pointed out that the urban effect in midlatitude areas is estimated to be 10%–20%. The solid urban signal in tropical cities is attributed to the local nature of rainfall in the tropical atmosphere that gives more chance for the urban effect to stand out in the rainfall episode, especially in the case a city is surrounded by nature vegetation land cover (Doan et al. 2021).
The flow of physical processes underlying the rainfall change is illustrated in Fig. 15. The starting point of the urban effect lies in the modification of surface land use/cover that leads to the modification of surface heat balance in terms of sensible heat increase through the reduction of evapotranspiration. The increased sensible heating enhances the UHI effect and destabilizes the low-level atmosphere in terms of invigorating boundary layer mixing. This leads to enhanced horizontal convergence of air moisture, which is pointed out in previous literature (e.g., Rozoff et al. 2003). Because of the city’s geographical location close to the sea, the UHI effect leads to a broadening land/sea temperature contrast, which encourages stronger sea breeze to develop and bring more moisture to converge over the urban area. This interaction between urban and sea breeze has been addressed in previous studies (Argüeso et al. 2016; Doan et al. 2021) for tropical coastal cities. The novelty of this study is that, for the first time, we are employing the current theories for the African megacity of Lagos.
Conceptual map of urban impact on precipitation over Lagos. Variables are sensible heat flux (QH), surface skin temperature (Tsfc), surface air temperature (T2m), horizontal wind speed (Wspd) (positive value for landward direction), planetary boundary layer height (PBLH), surface air pressure (Psfc), water vapor flux (WVF) (positive value for landward direction), and precipitation (Prcp).
Citation: Journal of Applied Meteorology and Climatology 62, 2; 10.1175/JAMC-D-22-0048.1
4. Conclusions
This study sought to be the first to investigate the urban effect on extreme rainfall events in Lagos, Nigeria, and hoped to fill the gap in the current body of knowledge about such processes on the African continent. We use the convection-permitting WRF Model within an ensemble framework using different physical schemes and boundary/initial conditions, to simulate and track the urban footprint on extreme rainfall of the selected event, 30 May 2006. The ensemble framework is chosen to identify the best model configuration and reduce the uncertainty in rainfall modeling in order to deliver robust results for this discussion. The simulated results are evaluated against the satellite rainfall product CMORPH, and the best four of the ensemble members are used for mechanism analysis.
Comparisons between the control run (the best four with status-quo land condition) and the nonurban run (four simulations corresponding to the best four control run’s configurations, with the urban area replaced with forest) show that the urban area enhances the extreme rainfall event. The control run produces more rainfall over the city by 20%–30% than the nonurban counterpart. This enhancement is attributed to the urban influence of reduced evapotranspiration, resulting in an increase of sensible heat flux (by 75 W m−2). This, in turn, increases the urban heat island effect (+1.5 K of air surface temperature increase), instigating horizontal convergence, and thus, augmenting sea breeze. As a result, more moisture is transported from the southern sea area to converge over the city, creating favorable conditions for convection and rainfall processes.
The results of our study offer broader implications to the forecasting and modeling field in terms of raising awareness of the urban-associated extreme rainfall change in an African megacity. We believe this study is only the first step toward more robust and beneficial identification of urban hazards in the region. This and future studies on this order are essential to the imperatives for achieving the U.N. Sustainable Development Goals.
The study’s results have answered the questions posed at the beginning of this study, that is, does the urban area modify the extreme precipitation occurring over an African megacity, Lagos? What physical processes are responsible for these changes? However, our results are limited to a case study and specific settings with regard to numerical modeling. We suggest that more research investigating the subject from different aspects, including climatological and observational analysis, is necessary to seek more conclusive and robust answers to the questions.
Acknowledgments.
This research was supported by JSPS KAKENHI Grants 20K13258 and 19H01155 and the Multidisciplinary Cooperative Research Program in the Center for Computational Sciences, University of Tsukuba, NASA IDS Grant 80NSSC20K1262, and USDA NIFA Grant 2015-67003-23460. The National Center for Atmospheric Research is sponsored by the National Science Foundation. The authors thank Karen Slater (NCAR RAL) for her help in improving the writing of this paper.
Data availability statement.
The Weather Research and Forecasting Model used to conduct numerical simulations in this study is openly available from UCAR’s WRF Users Page (https://www2.mmm.ucar.edu/wrf/users/download/get_sources.html). The satellite rainfall product CMORPH used for simulation verification in this study is available from NOAA’s National Weather Service Climate Prediction Center (https://www.cpc.ncep.noaa.gov/products/janowiak/cmorph_description.html). The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request.
APPENDIX
Additional Figures for 16 Ensemble Members
Figure A1 gives the spatial distribution of precipitation when it reaches the peak over the Lagos region, as in Fig. 3, but expanded to show simulated results from 16 ensemble members. Figure A2 gives the urban footprint on that spatial distribution of precipitation, as in Fig. 4, but expanded to show simulated results from 16 ensemble members.
Distribution of precipitation when it reaches the peak over the Lagos region: Simulated results from 16 ensemble members are shown. Among them, members 3, 4, 7, and 14 are considered to be best performing (in comparing with CMORPH) and are then used for analysis. Note that 1700 UTC indicates accumulated precipitation for 1500–1700 UTC.
Citation: Journal of Applied Meteorology and Climatology 62, 2; 10.1175/JAMC-D-22-0048.1
Urban footprint on spatial distribution of rainfall event at its peak (1500–1700 UTC): Spatial distribution of peak time (16-member ensemble mean) from (a) CTRL and (b) NoURB. (c) The difference between CTRL [in (a)] and NoURB [in (b)]. (d) Time series of the spatial average over Lagos.
Citation: Journal of Applied Meteorology and Climatology 62, 2; 10.1175/JAMC-D-22-0048.1
REFERENCES
Aliu, I. R., 2020: Energy efficiency in postpaid-prepaid metered homes: Analyzing effects of socio-economic, housing, and metering factors in Lagos, Nigeria. Energy Effic., 13, 853–869, https://doi.org/10.1007/s12053-020-09850-y.
Argüeso, D., A. Di Luca, and J. P. Evans, 2016: Precipitation over urban areas in the western Maritime Continent using a convection-permitting model. Climate Dyn., 47, 1143–1159, https://doi.org/10.1007/s00382-015-2893-6.
Baik, J.-J., Y.-H. Kim, J.-J. Kim, and J.-Y. Han, 2007: Effects of boundary-layer stability on urban heat island-induced circulation. Theor. Appl. Climatol., 89, 73–81, https://doi.org/10.1007/s00704-006-0254-4.
Bassett, R., P. J. Young, G. S. Blair, F. Samreen, and W. Simm, 2020: The megacity Lagos and three decades of urban heat island growth. J. Appl. Meteor. Climatol., 59, 2041–2055, https://doi.org/10.1175/JAMC-D-20-0059.1.
Burian, S. J., and J. M. Shepherd, 2005: Effect of urbanization on the diurnal rainfall pattern in Houston. Hydrol. Processes, 19, 1089–1103, https://doi.org/10.1002/hyp.5647.
Changnon, S. A., 1981: Midwestern cloud, sunshine and temperature trends since 1901: Possible evidence of jet contrail effects. J. Appl. Meteor. Climatol., 20, 496–508, https://doi.org/10.1175/1520-0450(1981)020<0496:MCSATT>2.0.CO;2.
Chen, F., and Coauthors, 2011: The integrated WRF/urban modelling system: Development, evaluation, and applications to urban environmental problems. Int. J. Climatol., 31, 273–288, https://doi.org/10.1002/joc.2158.
Dee, D. P., and Coauthors, 2011: The ERA-Interim reanalysis: Configuration and performance of the data assimilation system. Quart. J. Roy. Meteor. Soc., 137, 553–597, https://doi.org/10.1002/qj.828.
Diedhiou, A., S. Janicot, A. Viltard, and P. de Felice, 1998: Evidence of two regimes of easterly waves over West Africa and tropical Atlantic. Geophys. Res. Lett., 25, 2805–2808, https://doi.org/10.1029/98GL02152.
Doan, Q.-V., and H. Kusaka, 2016: Numerical study on regional climate change due to the rapid urbanization of greater Ho Chi Minh City’s metropolitan area over the past 20 years. Int. J. Climatol., 36, 3633–3650, https://doi.org/10.1002/joc.4582.
Doan, Q.-V., A. Dipankar, A. Simón-Moral, C. Sanchez, V. Prasanna, M. Roth, and X.-Y. Huang, 2021: Urban-induced modifications to the diurnal cycle of rainfall over a tropical city. Quart. J. Roy. Meteor. Soc., 147, 1189–1201, https://doi.org/10.1002/qj.3966.
Doan, V. Q., H. Kusaka, and T. M. Nguyen, 2019: Roles of past, present, and future land use and anthropogenic heat release changes on urban heat island effects in Hanoi, Vietnam: Numerical experiments with a regional climate model. Sustainable Cities Soc., 47, 101479, https://doi.org/10.1016/j.scs.2019.101479.
Ezema, I. C., A. O. Olotuah, and O. I. Fagbenle, 2016: Evaluation of energy use in public housing in Lagos, Nigeria: Prospects for renewable energy sources. Int. J. Renewable Energy Dev., 5, 15–24, https://doi.org/10.14710/ijred.5.1.15-24.
Faisal Koko, A., W. Yue, G. Abdullahi Abubakar, R. Hamed, and A. A. Noman Alabsi, 2021: Analyzing urban growth and land cover change scenario in Lagos, Nigeria using multi-temporal remote sensing data and GIS to mitigate flooding. Geomatics Nat. Hazards Risk, 12, 631–652, https://doi.org/10.1080/19475705.2021.1887940.
Freitag, B. M., U. S. Nair, and D. Niyogi, 2018: Urban modification of convection and rainfall in complex terrain. Geophys. Res. Lett., 45, 2507–2515, https://doi.org/10.1002/2017GL076834.
Gero, A. F., and A. J. Pitman, 2006: The impact of land cover change on a simulated storm event in the Sydney basin. J. Appl. Meteor. Climatol., 45, 283–300, https://doi.org/10.1175/JAM2337.1.
Golroudbary, V. R., Y. Zeng, C. M. Mannaerts, and Z. B. Su, 2017: Detecting the effect of urban land use on extreme precipitation in the Netherlands. Wea. Climate Extremes, 17, 36–46, https://doi.org/10.1016/j.wace.2017.07.003.
Grady Dixon, P., and T. L. Mote, 2003: Patterns and causes of Atlanta’s urban heat island-initiated precipitation. J. Appl. Meteor. Climatol., 42, 1273–1284, https://doi.org/10.1175/1520-0450(2003)042<1273:PACOAU>2.0.CO;2.
Gunwani, P., and M. Mohan, 2017: Sensitivity of WRF model estimates to various PBL parameterizations in different climatic zones over India. Atmos. Res., 194, 43–65, https://doi.org/10.1016/j.atmosres.2017.04.026.
Hamdi, R., D. Degrauwe, and P. Termonia, 2012: Coupling the Town Energy Balance (TEB) scheme to an operational limited-area NWP model: Evaluation for a highly urbanized area in Belgium. Wea. Forecasting, 27, 323–344, https://doi.org/10.1175/WAF-D-11-00064.1.
Hong, S.-Y., and J.-O. J. Lim, 2006: The WRF single-moment 6-class microphysics scheme (WSM6). J. Korean Meteor. Soc., 42, 129–151.
Hong, S.-Y., Y. Noh, and J. Dudhia, 2006: A new vertical diffusion package with an explicit treatment of entrainment processes. Mon. Wea. Rev., 134, 2318–2341, https://doi.org/10.1175/MWR3199.1.
Hoornweg, D., and K. Pope, 2017: Population predictions for the world’s largest cities in the 21st century. Environ. Urban., 29, 195–216, https://doi.org/10.1177/0956247816663557.
Horton, R. E., 1921: Thunderstorm-breeding spots. Mon. Wea. Rev., 49, 193–194, https://doi.org/10.1175/1520-0493(1921)49<193a:TS>2.0.CO;2.
Hourdin, F., and Coauthors, 2010: AMMA-Model Intercomparison Project. Bull. Amer. Meteor. Soc., 91, 95–104, https://doi.org/10.1175/2009BAMS2791.1.
Huff, F. A., and S. A. Changnon Jr., 1972: Climatological assessment of urban effects on precipitation as St. Louis. J. Appl. Meteor. Climatol., 11, 823–842, https://doi.org/10.1175/1520-0450(1972)011<0823:CAOUEO>2.0.CO;2.
IPCC, 2023: AR6 Synthesis Report: Climate Change 2023. Cambridge University Press, in press.
Jee, J.-B., and S. Kim, 2017: Sensitivity study on high-resolution WRF precipitation forecast for a heavy rainfall event. Atmosphere, 8, 96, https://doi.org/10.3390/atmos8060096.
Joyce, R. J., J. E. Janowiak, P. A. Arkin, and P. Xie, 2004: CMORPH: A method that produces global precipitation estimates from passive microwave and infrared data at high spatial and temporal resolution. J. Hydrometeor., 5, 487–503, https://doi.org/10.1175/1525-7541(2004)005<0487:CAMTPG>2.0.CO;2.
Kaufmann, R. K., K. C. Seto, A. Schneider, Z. Liu, L. Zhou, and W. Wang, 2007: Climate response to rapid urban growth: Evidence of a human-induced precipitation deficit. J. Climate, 20, 2299–2306, https://doi.org/10.1175/JCLI4109.1.
Kusaka, H., and F. Kimura, 2004: Coupling a single-layer urban canopy model with a simple atmospheric model: Impact on urban heat island simulation for an idealized case. J. Meteor. Soc. Japan, 82, 67–80, https://doi.org/10.2151/jmsj.82.67.
Kusaka, H., F. Kimura, H. Hirakuchi, and M. Mizutori, 2000: The effects of land-use alteration on the sea breeze and daytime heat island in the Tokyo metropolitan area. J. Meteor. Soc. Japan, 78, 405–420, https://doi.org/10.2151/jmsj1965.78.4_405.
Kusaka, H., H. Kondo, Y. Kikegawa, and F. Kimura, 2001: A simple single-layer urban canopy model for atmospheric models: Comparison with multi-layer and slab models. Bound.-Layer Meteor., 101, 329–358, https://doi.org/10.1023/A:1019207923078.
Kusaka, H., K. Nawata, A. Suzuki-Parker, Y. Takane, and N. Furuhashi, 2014: Mechanism of precipitation increase with urbanization in Tokyo as revealed by ensemble climate simulations. J. Appl. Meteor. Climatol., 53, 824–839, https://doi.org/10.1175/JAMC-D-13-065.1.
Kusaka, H., A. Nishi, M. Mizunari, and H. Yokoyama, 2019: Urban impacts on the spatiotemporal pattern of short-duration convective precipitation in a coastal city adjacent to a mountain range. Quart. J. Roy. Meteor. Soc., 145, 2237–2254, https://doi.org/10.1002/qj.3555.
Li, W., and Coauthors, 2019: Annual precipitation and daily extreme precipitation distribution: Possible trends from 1960 to 2010 in urban areas of China. Geomatics Nat. Hazards Risk, 10, 1694–1711, https://doi.org/10.1080/19475705.2019.1609604.
Li, Y., and Coauthors, 2020: Strong intensification of hourly rainfall extremes by urbanization. Geophys. Res. Lett., 47, e2020GL088758, https://doi.org/10.1029/2020GL088758.
Lim, K.-S. S., and S.-Y. Hong, 2010: Development of an effective double-moment cloud microphysics scheme with prognostic cloud condensation nuclei (CCN) for weather and climate models. Mon. Wea. Rev., 138, 1587–1612, https://doi.org/10.1175/2009MWR2968.1.
Lin, C.-Y., W.-C. Chen, P.-L. Chang, and Y.-F. Sheng, 2011: Impact of the urban heat island effect on precipitation over a complex geographic environment in northern Taiwan. J. Appl. Meteor. Climatol., 50, 339–353, https://doi.org/10.1175/2010JAMC2504.1.
Liu, J., and D. Niyogi, 2019: Meta-analysis of urbanization impact on rainfall modification. Sci. Rep., 9, 7301, https://doi.org/10.1038/s41598-019-42494-2.
Luong, T. M., H. P. Dasari, and I. Hoteit, 2020: Impact of urbanization on the simulation of extreme rainfall in the City of Jeddah, Saudi Arabia. J. Appl. Meteor. Climatol., 59, 953–971, https://doi.org/10.1175/JAMC-D-19-0257.1.
Mitra, C., J. M. Shepherd, and T. Jordan, 2012: On the relationship between the premonsoonal rainfall climatology and urban land cover dynamics in Kolkata city, India. Int. J. Climatol., 32, 1443–1454, https://doi.org/10.1002/joc.2366.
Mlawer, E. J., S. J. Taubman, P. D. Brown, M. J. Iacono, and S. A. Clough, 1997: Radiative transfer for inhomogeneous atmospheres: RRTM, a validated correlated-k model for the longwave. J. Geophys. Res., 102, 16 663–16 682, https://doi.org/10.1029/97JD00237.
Mote, T. L., M. C. Lacke, and J. M. Shepherd, 2007: Radar signatures of the urban effect on precipitation distribution: A case study for Atlanta, Georgia. Geophys. Res. Lett., 34, L20710, https://doi.org/10.1029/2007GL031903.
Niyogi, D., P. Pyle, M. Lei, S. P. Arya, C. M. Kishtawal, M. Shepherd, F. Chen, and B. Wolfe, 2011: Urban modification of thunderstorms: An observational storm climatology and model case study for the Indianapolis urban region. J. Appl. Meteor. Climatol., 50, 1129–1144, https://doi.org/10.1175/2010JAMC1836.1.
Ojeh, V. N., A. A. Balogun, and A. A. Okhimamhe, 2016: Urban-rural temperature differences in Lagos. Climate, 4, 29, https://doi.org/10.3390/cli4020029.
Oke, T. R., G. Mills, A. Christen, and J. A. Voogt, 2017: Urban Climates. Cambridge University Press, 546 pp.
Ooi, M. C. G., A. Chan, K. Subramaniam, K. I. Morris, and M. Y. Oozeer, 2017: Interaction of urban heating and local winds during the calm intermonsoon seasons in the tropics. J. Geophys. Res. Atmos., 122, 11 499–11 523, https://doi.org/10.1002/2017JD026690.
Orville, R. E., and Coauthors, 2001: Enhancement of cloud-to-ground lightning over Houston, Texas. Geophys. Res. Lett., 28, 2597–2600, https://doi.org/10.1029/2001GL012990.
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.
Peel, M. C., B. L. Finlayson, and T. A. McMahon, 2007: Updated world map of the Köppen-Geiger climate classification. Hydrol. Earth Syst. Sci., 11, 1633–1644, https://doi.org/10.5194/hess-11-1633-2007.
Pleim, J. E., 2007: A combined local and nonlocal closure model for the atmospheric boundary layer. Part I: Model description and testing. J. Appl. Meteor. Climatol., 46, 1383–1395, https://doi.org/10.1175/JAM2539.1.
Qian, Y., and Coauthors, 2022: Urbanization impact on regional climate and extreme weather: Current understanding, uncertainties, and future research directions. Adv. Atmos. Sci., 39, 819–860, https://doi.org/10.1007/s00376-021-1371-9.
Rozoff, C. M., W. R. Cotton, and J. O. Adegoke, 2003: Simulation of St. Louis, Missouri, land use impacts on thunderstorms. J. Appl. Meteor. Climatol., 42, 716–738, https://doi.org/10.1175/1520-0450(2003)042<0716:SOSLML>2.0.CO;2.
Sailor, D. J., M. Georgescu, J. M. Milne, and M. A. Hart, 2015: Development of a national anthropogenic heating database with an extrapolation for international cities. Atmos. Environ., 118, 7–18, https://doi.org/10.1016/j.atmosenv.2015.07.016.
Shastri, H., S. Paul, S. Ghosh, and S. Karmakar, 2015: Impacts of urbanization on Indian summer monsoon rainfall extremes. J. Geophys. Res. Atmos., 120, 496–516, https://doi.org/10.1002/2014JD022061.
Shem, W., and M. Shepherd, 2009: On the impact of urbanization on summertime thunderstorms in Atlanta: Two numerical model case studies. Atmos. Res., 92, 172–189, https://doi.org/10.1016/j.atmosres.2008.09.013.
Shin, H. H., and S.-Y. Hong, 2011: Intercomparison of planetary boundary-layer parametrizations in the WRF model for a single day from CASES-99. Bound.-Layer Meteor., 139, 261–281, https://doi.org/10.1007/s10546-010-9583-z.
Simón-Moral, A., A. Dipankar, Q.-V. Doan, C. Sanchez, M. Roth, E. Becker, and X.-Y. Huang, 2021: Urban intensification of convective rainfall over the Singapore–Johor Bahru region. Quart. J. Roy. Meteor. Soc., 147, 3665–3680, https://doi.org/10.1002/qj.4147.
Skamarock, W. C., and Coauthors, 2008: A description of the Advanced Research WRF version 3. NCAR Tech. Note NCAR/TN-475+STR, 113 pp., https://doi.org/10.5065/D68S4MVH.
Song, H.-J., and B.-J. Sohn, 2018: An evaluation of WRF microphysics schemes for simulating the warm-type heavy rain over the Korean Peninsula. Asia-Pac. J. Atmos. Sci., 54, 225–236, https://doi.org/10.1007/s13143-018-0006-2.
Sylla, M. B., I. Diallo, and J. S. Pal, 2013: West African monsoon in state-of-the-science regional climate models. Climate Variability: Regional and Thematic Patterns, A. Tarhule, Ed., IntechOpen, 3–36, https://doi.org/10.5772/55140.
Tewari, M., F. Chen, J. Dudhia, P. Ray, S. Miao, E. Nikolopoulos, and L. Treinish, 2021: Understanding the sensitivity of WRF hindcast of Beijing extreme rainfall of 21 July 2012 to microphysics and model initial time. Atmos. Res., 271, 106085, https://doi.org/10.1016/j.atmosres.2022.106085.
Thielen, J., W. Wobrock, A. Gadian, P. G. Mestayer, and J.-D. Creutin, 2000: The possible influence of urban surfaces on rainfall development: A sensitivity study in 2D in the meso-γ-scale. Atmos. Res., 54, 15–39, https://doi.org/10.1016/S0169-8095(00)00041-7.
Tiedtke, M., 1989: A comprehensive mass flux scheme for cumulus parameterization in large-scale models. Mon. Wea. Rev., 117, 1779–1800, https://doi.org/10.1175/1520-0493(1989)117<1779:ACMFSF>2.0.CO;2.
United Nations, 2015: Sustainable Development Goals. Department of Economic and Social Affairs, https://sdgs.un.org/goals.
United Nations, 2018: 2018 revision of world urbanization prospects. Department of Economic and Social Affairs, https://www.un.org/development/desa/publications/2018-revision-of-world-urbanization-prospects.html.
Wang, D., P. Jiang, G. Wang, and D. Wang, 2015: Urban extent enhances extreme precipitation over the Pearl River Delta, China. Atmos. Sci. Lett., 16, 310–317, https://doi.org/10.1002/asl2.559.
Wang, J., F. Chen, Q.-V. Doan, and Y. Xu, 2021: Exploring the effect of urbanization on hourly extreme rainfall over Yangtze River Delta of China. Urban Climate, 36, 100781, https://doi.org/10.1016/j.uclim.2021.100781.
Westcott, N. E., 1995: Summertime cloud-to-ground lightning activity around major midwestern urban areas. J. Appl. Meteor. Climatol., 34, 1633–1642, https://doi.org/10.1175/1520-0450-34.7.1633.
Xie, P., R. Joyce, S. Wu, S.-H. Yoo, Y. Yarosh, F. Sun, and R. Lin, 2019: NOAA Climate Data Record (CDR) of CPC Morphing Technique (CMORPH) High Resolution Global Precipitation Estimates, Version 1. NOAA National Centers for Environmental Information, accessed 14 April 2022, https://doi.org/10.25921/w9va-q159.
Yonetani, T., 1982: Increase in number of days with heavy precipitation in Tokyo urban area. J. Appl. Meteor. Climatol., 21, 1466–1471, https://doi.org/10.1175/1520-0450(1982)021<1466:IINODW>2.0.CO;2.
Zhang, C. L., F. Chen, S. G. Miao, Q. C. Li, X. A. Xia, and C. Y. Xuan, 2009: Impacts of urban expansion and future green planting on summer precipitation in the Beijing metropolitan area. J. Geophys. Res., 114, D02116, https://doi.org/10.1029/2008JD010328.
Zhang, Y., and Coauthors, 2019: Regional patterns of extreme precipitation and urban signatures in metropolitan areas. J. Geophys. Res. Atmos., 124, 641–663, https://doi.org/10.1029/2018JD029718.
Zhong, S., and X.-Q. Yang, 2015: Ensemble simulations of the urban effect on a summer rainfall event in the Great Beijing metropolitan area. Atmos. Res., 153, 318–334, https://doi.org/10.1016/j.atmosres.2014.09.005.
