Skip to main content
  • Loading metrics

Impacts of projected urban growth on simulated near-surface temperature in Mexico City Metropolitan Area: Implications for urban vulnerability

  • Yosune Miquelajauregui,

    Roles Conceptualization, Data curation, Formal analysis, Methodology, Writing – original draft, Writing – review & editing

    Affiliation Laboratorio Nacional de Ciencias de la Sostenibilidad, Instituto de Ecologia, Universidad Nacional Autónoma de México, Mexico City, Mexico

  • Erika Danaé López-Espinoza ,

    Roles Data curation, Formal analysis, Investigation, Methodology, Writing – original draft, Writing – review & editing

    Affiliation Instituto de Ciencias de la Atmosfera y Cambio Climatico, Universidad Nacional Autónoma de México, Mexico City, Mexico

  • Erika Luna Pérez,

    Roles Data curation, Formal analysis, Methodology, Writing – review & editing

    Affiliation Laboratorio Nacional de Ciencias de la Sostenibilidad, Instituto de Ecologia, Universidad Nacional Autónoma de México, Mexico City, Mexico

  • Paola Gómez-Priego,

    Roles Methodology, Writing – original draft, Writing – review & editing

    Affiliation Laboratorio Nacional de Ciencias de la Sostenibilidad, Instituto de Ecologia, Universidad Nacional Autónoma de México, Mexico City, Mexico

  • Luis A. Bojórquez-Tapia,

    Roles Conceptualization, Funding acquisition, Methodology, Writing – original draft, Writing – review & editing

    Affiliation Laboratorio Nacional de Ciencias de la Sostenibilidad, Instituto de Ecologia, Universidad Nacional Autónoma de México, Mexico City, Mexico

  • Lourdes P. Aquino Martínez,

    Roles Writing – original draft, Writing – review & editing

    Affiliation Escuela Nacional de Ciencias de la Tierra, Universidad Nacional Autónoma de México, Mexico City, Mexico

  • Arturo I. Quintanar

    Roles Writing – original draft, Writing – review & editing

    Affiliation Instituto de Ciencias de la Atmosfera y Cambio Climatico, Universidad Nacional Autónoma de México, Mexico City, Mexico


Urbanization impacts the surface temperature fields increasing the vulnerability of urban residents to heat exposure. Identifying vulnerable urban populations to extreme heat exposure is crucial to develop mitigation and adaptation strategies towards sustainability. We used an urban growth model (SLEUTH) to simulate emerging urban areas in Mexico City Metropolitan Area (MCMA) under a hypothetical land-use policy scenario projected to 2060 in which no restrictions were posed to urban growth. SLEUTH outputs were used in the numerical model Weather Research and Forecasting (WRF) to quantify expected changes in near-surface temperature within the MCMA. We calculated and mapped heat exposure as differences in average (Tmean), maximum (Tmax) and minimum (Tmin) temperatures over the diurnal cycle between future and current land cover conditions. Population vulnerability to projected increases in heat exposure was determined using a set of socioeconomic indicators. SLEUTH simulations showed an urban area expansion of nearly 4,790 km2 by 2060. Overall, changes in Tmin were greater than changes observed for Tmax and Tmean. Tmean, Tmax and Tmin increases up to 0.6°C, 1.3°C and 2.6°C, respectively, were recorded for the MCMA with greatest temperature changes observed in the State of Mexico. Results suggested the presence of socioeconomic disparities in the projected spatial exposure of urban-induced heat in MCMA. We argue that our results could be used to inform and guide locally tailored actions aimed at reducing exposure and increasing population´s capacities to cope and adapt to future threats.

1 Introduction

Urbanization is one of the main driving forces of environmental change across multiple scales [1, 2]. The rate of urbanization is increasing rapidly around the world expecting urban land cover to nearly tripling by 2030 [3]. Urbanization increases population density and leads to the replacement of natural land cover with built structures of greater impervious surface areas [4]. Consequently, urbanization can alter the biogeophysical and biogeochemical variables that determine the local and regional climate which impacts can be observed in changes in temperature and precipitation patterns [57]. An increase of approximately 0.23°C has been attributed to urban expansion across the continental United States over the past half-century [8]. Similarly, the escalating temperatures attributed to urbanization have been documented in Asian megacities, revealing temperature increases ranging from 0.74°C to 1.6°C, particularly in the densest areas of the city during nighttime [7, 911].

This urbanization-driven temperature rise intensifies the vulnerability of urban residents to heat exposure [9, 12]. Additionally, climate change is likely to exacerbate these risks as severe heat waves and droughts become more frequent [13]. Significantly, instances of extreme heat exposure have become more prevalent, affecting half of global urban settlements, and impacting nearly a quarter of the world´s population since the end of the 20th century [14]. Consequently, the identification of high-risk urban populations vulnerable to extreme heat exposure is fundamental for promoting sustainable urban development.

Vulnerability may be characterized as the susceptibility of urban populations to harm from exposure to urban heat [15, 16]. Nevertheless, the uneven distribution of vulnerability is evident among diverse populations [1719]. Disparities in vulnerability are related to the interplay of socioeconomic, demographic, infrastructural and environmental determinants [20, 21]. Hence, this underscores the imperative to craft site-specific mitigation and adaptation strategies targeted at diminishing exposure and amplifying capacities to cope and adapt to future threats [22, 23]. In pursuit of this objective, analytical and modelling tools have become indispensable for capturing, quantifying, and visually representing the spatial distribution of heat exposure within cities [11, 24, 25].

The exploration of urban climate has traditionally centered on the analysis of urban heat patterns often employing numerical models such as MUKLIMO_3 [26], UrbClim [27], and MESO-NH [28]. However, simulations have only encompassed either specific sections of a city for several months or entire cities spanning decades [2931]. Notably, the Weather Research and Forecasting (WRF) model stands as an exceptionally versatile numerical weather prediction and atmospheric simulation system, extensively utilized across various applications, including urban climate studies [11, 3234]. In particular, the integration of the WRF model with urban sprawl simulation tools, like SLEUTH [35], confers the capability to generate fine-grained weather forecasts while simultaneously considering urbanization patterns at both local and regional scales [9]. This stands as perhaps the most important consideration for mitigating the vulnerability of urban populations, particularly in cities located in the Global South, such as the Mexico City Metropolitan Area (MCMA).

Many of the climate studies based on WRF modeling carried out in the MCMA, have primarily focused on investigating the impact of urbanization on regional meteorology. For instance, [36] studied the sensitivity of temperature forecasts to changes in urban coverage. Similarly, [37], undertook an analysis to ascertain the spatiotemporal variations in near-surface temperature and precipitation resulting from the current urban landscape when compared to pre-settlement conditions. More recently, [38, 39] explored alterations in the diurnal cycles of temperature, precipitation and wind fields associated with urban sprawl. Despite the notable progress of the studies, there remains a gap in urban climate research pertaining to local and long-term exposure to extreme heat, particularly considering the inherent uncertainty of future urban expansion in the MCMA.

In this study we integrated WRF with SLEUTH to examine the vulnerability of urban populations to future heat exposure risks induced by urban sprawl in Mexico City Metropolitan Area under a no restrictive urban growth scenario. We coupled SLEUTH outputs with WRF to simulate expected changes in near-surface temperature within the MCMA. We calculated the Gower’s residuals to assess the vulnerability of urban residents to projected heat exposure using socioeconomic indicators derived from the 2020 national census. We argue that the results of this study contribute to identify potential vulnerable areas to heat exposure, which could consequently lead to unequal distribution of vulnerability. Although outside the scope of this study, our results can also be combined with climate change projections to assess synergistic impacts on urban vulnerability to heat exposure. In addition, the resulting insights of this research can contribute to the understanding of the overall quality of life within urban environments the MCMA. The SLEUTH/WRF modelling framework presented here possesses the potential to furnish decision-makers with valuable inputs, guiding the formulation of strategies and initiatives aimed at mitigating vulnerability to extreme heat.

2 Methods

2.1 Study area

The Mexico City Metropolitan Area is one of the world´s five largest cities with nearly 22 million inhabitants [40] spreading over an urban area of 7,954 km2 (Fig 1A; [41]). The MCMA is situated in the Valley of Mexico, which is surrounded by mountain ranges on its southern, western, and eastern sides with an opening to the Mexican Plateau to the north and a mountain gap to the southeast [42]. The MCMA is located in a high, elevated lacustrine plain (2240 m.a.s.l) which once supported a large system of interconnected lakes before being drained and transformed by urban landscapes ([22, 37]; S1 Fig). The climate in the MCMA is subhumid tropical tempered by altitude with a cold-dry season from November to February, a warm-dry season from March to April, and a rainy season from May to October. Annual temperatures average 17°C with maximum temperatures occurring during the warm-dry season when relative humidity is low [43]. The MCMA has been historically affected with heat waves, however, the frequency of days with maximum temperatures above 30°C has significantly increased since the second half of the 20th century [43]. During that period, the MCMA experienced a substantial population growth accompanied by high rates of land cover transformation and urban expansion that rapidly incorporated the formal and informal peripheries of the State of Mexico and Hidalgo [44, 45] (S2 Fig). The lack of access to public urban services and infrastructure in these urban peripheries has resulted in urban pockets characterized by marked socioeconomic inequalities [46]. According to urban planning information available from the Planning Committee for the Development of the State of Mexico (COPLADEM, for its Spanish acronym) and the National Institute for Statistic and Geography (Instituto Nacional de Estadística y Geografía, INEGI, for its Spanish acronym), the MCMA is composed of administrative units spanning three states. For this study, Mexico City, at the core of MCMA, was divided into sixteen administrative units, whereas the State of Mexico and Hidalgo, adjacent to Mexico City, into twenty and twelve units respectively (Fig 1A). Further analyses and results are reported at the administrative units’ level.

Fig 1. Study area and simulation domains.

a) The Mexico City Metropolitan Area including Mexico City (green), the State of Mexico (brown) and Hidalgo (lilac). Administrative units and associated IDs considered in this study are also shown (see Table 2). Panel b) shows the three WRF model computational domains. The blue rectangle corresponds to the MCMA (domain 3). States and municipalities shapefiles: Terms of use:

2.2 SLEUTH urban growth model

The SLEUTH cellular automaton urban growth model was developed by [35] and has been commonly applied to forecast urban growth in many regions of the world [4749]. The basic structure of SLEUTH consists of the inputs required for the model, the two sub-models within SLEUTH, the calibration and feedback processes, and the model prediction [50]. In SLEUTH, urban growth is modelled using a spatial two-dimensional grid of cells. Each cell in the study area has only two possible states: urbanized or non-urbanized. The change of state is regulated by predefined growth-rules applied on a cell-by-cell basis [50]. Potential cells for urbanization are selected at random and the growth-rules evaluate the conditions of the cell and its neighbors before a change in state can occur. SLEUTH uses four types of growth-rules to predict urban growth: spontaneous, new spreading center, edge, and road-influenced growth [47]. These growth-rules are applied sequentially during each iteration and controlled through the interaction of five growth coefficients which determine the probability of urbanization of a particular cell: dispersion, breed, spread, road gravity and slope [47, 50]. The values of each coefficient were obtained through a calibration procedure described in section 2.4.

2.3 Input datasets for SLEUTH

This study applied the SLEUTH 3.0 Beta model as an exploratory model obtained from the Project Gigalopolis digital repository [51]. SLEUTH requires as a minimum the following input data: slope, hillshade, four periods of urban extent maps, four periods of transportation maps and one exclusion map from urbanization zones. The inputs were derived from map layers created with the same projection and datum, spatial extent (MCMA), and resolution (1 ha) and were converted to grayscale GIF images. The slope layer expressed as percentage slope, as well as the hillshade layer were obtained from the digital elevation model [52]. The slope layer accounted for the influence of topography on urban growth, meaning that greater slope percentages would decrease the likelihood of urban development. The binary data for the urban-nonurban historical extent layers were obtained from INEGI for the years 1980, 1990, 2000 and 2014 [5355]. The binary data for the transportation layer (road-no road) for the year 2015 was obtained from INEGI [56]. The 2015 layer was the baseline to eliminate roads recursively and systematically for the years 2000, 1990, and 1980. The process of elimination consisted in analyzing the presence of roads through the visual inspection of Google Earth Imagery coupled with the analysis of the creation of new settlements that resulted in new roads. The exclusion from urbanization zone is a user-defined layer that allows the creation of alternative scenarios of urban growth restrictions associated to hypothetical land-use policies. The scenarios are built by assigning different levels of restriction or growth pressure to areas where urban growth is likely or not to occur in the future. That is, areas with high urban growth pressure have grid values of 0 indicating that they are 0% excluded from development. On the opposite, areas with low urban growth pressure have grid values of 100 indicating that they are 100% excluded from development. For this study, to illustrate a no restrictive urban growth scenario, the exclusion layer had values of 0 on al MCMA except for federal controlled areas such as airports, parks, as well as federal and locally fenced and policed areas, with values of 100. SLEUTH inputs are fully available and can be openly accessed through the following link (

2.4 SLEUTH calibration and urban growth forecast

Calibration is the most important step in modelling urban growth using SLEUTH [35]. The goal of this step is to derive a set of values for the growth coefficients that provide the best match between the modelled and observed patterns of urban growth for a given study area [50]. During the calibration process, SLEUTH tested for many combinations of the growth coefficients by performing multiple Monte Carlo runs from the starting (1980) to the last year (2014). Three calibration phases were performed as suggested by [47]: coarse, fine, and final. After each calibration phase, the range of the coefficients are narrowed based on the calibration results from the previous phase [50]. To evaluate the performance of different coefficient sets, we used the optimized SLEUTH metrics (OSM) developed by [57]. The OSM is a measurement of spatial fit between modelled and observed urban patterns and it is best used to eliminate redundancy among goodness of fit metrics while maintaining high accuracy in model predictions [50]. The OSM can take values from 0 to 1, 1 being the best fit given a set of coefficient values. Given the spatial heterogeneity in the determinants that influence of urban development in the MCMA such as economic, social, urban and enforcement of environmental policies, thirteen subregions within MCMA were delineated. These subregions were expected to result in better calibration coefficient values and produce more accurate simulated outputs. The average OSM across the 13 regions was 0.51 with a variance that ranged from 0.15 to 0.77. The best coefficient set was then used to explore the current urban trends without restrictions to forecast the probabilities of expansion to 2060. For these predictions 1000 Monte Carlo runs were performed and averaged to produce annual images of urban growth probabilities. The probabilities images were thresholded at 80% to create urban-nonurban binary images for the study area.

2.5 The Weather Research Forecast (WRF) modelling system

For the Valley of Mexico, efforts have been made to evaluate the performance and quality of WRF forecasts [38, 39, 58]. These evaluations have concluded that WRF reproduces, in an acceptable manner, the observed spatial and temporal meteorological patterns of the MCMA. Considering the above, we used the Advanced Research Version of WRF v. 4.0 to perform all atmospheric simulations. For this study, the nested grid configuration consisted of three domains (Fig 1B), an outer domain covering Central Mexico with a resolution of 15 km (domain 1), a nested domain that included the Valley of Mexico and the surrounding areas with a horizontal resolution of 5 km (domain 2), and a third domain (domain 3) covering the MCMA with a finer resolution of 1 km (Fig 1B). The three domains were configured with 35 terrain-following sigma levels in the vertical with the pressure at the model top set at 50 hPa. We used initial and boundary conditions from the Climate Forecast System reanalysis (CFSv2; [59]) every six hours with a horizontal resolution of approximately 0.2 degrees. We selected the interaction between the domains to be one-way on a Mercator projection. Throughout the simulation period, a Nudging Analysis was applied to temperature, the water vapor mixture ratio, and the components of the horizontal wind with a G coefficient of 3E-4, for each computational domain [60, 61]. Schemes used in the model physics were as follows: the Kain–Fritsch [62] for the cumulus parameterization, which was turned off for the domain 3; the Rapid Radiative Transfer Model for the long and short-wave radiation, and the Yonsei University scheme for the Planetary Boundary Layer [63]. The community Noah land surface model was used to represent land-surface processes [64].

2.6 Experimental setup and numerical experiments

We set up two numerical WRF experiments representing current and future urban land conditions. The current urban land conditions used in the control experiment (hereafter “CNTRL”, Fig 2—left) were derived from the 2014 land cover data produced by INEGI serie VI (see, Accessed on 30 March 2022). For the projected experiment (hereafter “U2060”), we used future urban growth conditions projected to 2060 under a no restrictive urban growth scenario derived from SLEUTH outputs (Fig 2—right). To ensure consistency between the CNTRL and U2060 experiments, both land cover maps were converted to the 24 classes classification scheme proposed by the USGS [64]. Empirical evidence indicates that temperatures during neutral years in the Valley of Mexico can typically exhibit up to 1°C above long-term average. In contrast, during ENSO years, near-surface temperatures have been observed to exceed this threshold [37]. To mitigate the potential impact of such large-scale anomalies on local circulations within the MCMA basin, we selected a simulation year (2011) characterized by neutral conditions (-0.4 °C) of the ENSO (El Niño-Southern Oscillation). Simulation experiments were initiated on 26 May 2011 and executed until 31 June of 2011, with output written every hour through the end of the simulation period. The initial six days of the two sets of simulations (from May 26 to May 31) were discarded and considered as model spin-up.

Fig 2. Simulated urban growth in the MCMA.

Left: Urban land cover (in red) for the CNTRL experiment (current LULC), Right: U2060 (projected to 2060) experiments. Different colors represent land cover categories according to the USGS scheme. The white dotted line represents the selected west-east transect at 19.8°N (Fig 4). Land use and land cover shapefile: States shapefile: Terms of use:

2.7 WRF validation

To assess the performance of the WRF model, we conducted a validation process that entailed comparing observational data with WRF-generated outputs. Historical near-surface temperature data was collected from 21 automatic weather stations managed by the Servicio Meteorológico Nacional Mexicano (SMN) (S3 Fig). Observed temperature records for the baseline CNTRL simulation period were collected at 10-minute intervals from which we derived hourly averages. We calculated the root mean square error (RMSE) and the mean absolute error (MAE) by comparing hourly observed and simulated temperatures records for each meteorological station (Table 1). The WRF model successfully replicated the variability of observed temperatures within the MCMA based on calculated RMSE and MAE estimates (Table 1).

Table 1. Selected monitoring weather stations from the Servicio Meteorológico Nacional used to perform WRF model validation.

Basic information for each station including station name, ID, coordinates, observed and simulated (mean±sd) temperatures and mean absolute errors (MAE) and root mean square errors (RMSE) are also shown.

2.8 Impacts of projected urban growth on near-surface temperature

To evaluate impacts of projected urban growth on near-surface temperature patterns within MCMA we calculated average (Tmean) maximum (Tmax) and minimum (Tmin) temperatures changes over the diurnal cycle between the U2060 and the CNTRL experiments from domain 3 outputs. Variables such as relative humidity, wind speed, and mean radiant temperature were excluded from the analysis [65]. We mapped Tmean, Tmax and Tmin changes at the administrative unit´s level and municipalities over the MCMA. We also selected a west-east transect at 19.8° N (see Fig 2, white dotted line) to determine the relative contribution of different land cover conversion on projected temperature changes. The selected transect included the representative land cover categories found in the MCMA according to the classification scheme proposed by the USGS (Fig 2; [66]).

2.9 Assessment of socioeconomic vulnerability

We assessed the socioeconomic vulnerability within MCMA to projected changes in near-surface temperature using indicators for population, health, economy, and infrastructure, which themselves are associated to key vulnerability drivers (see [67], S1 Table). Population indicators included the total population (pop) and the population older than 60 years of age (pop60 >). Health and economic indicators included the total population without access to public health services (pop_nomed) and the unemployed population (pop_nowork), respectively. Infrastructure indicators included the number of households without electricity (h_noelec), without access to drinking water (h_nwater), without drainage (h_nosewage) and the number of households with soil floors (h_soil). Data for the indicators were obtained from the 2020 national census [68] aggregating them at the administrative unit’s level (Fig 1A; data can be accessed through We calculated the Gower’s residuals [6971] to numerically assess the relative vulnerability or sensitivity of the administrative units to each indicator. The Gower’s residuals are a double centering technique commonly used in spatial analysis to evaluate the deviation of individual units from the overall pattern observed in the data. The procedure consists in adjusting a matrix Z containing the normalized values for each socioeconomic indicator within a respective administrative unit. The matrix Z is adjusted by applying Eq 1: (1) where vgj is the Gower’s residuals, zgj is the normalized value, zg. is the indicator´s mean for each administrative unit, z.j is the mean of socioeconomic indicators, z‥ is the mean for the whole matrix, and g and j are indices associated to each administrative unit and socioeconomic indicator, respectively. The results capture the differences between administrative units that exhibit higher or lower vulnerability compared to the average vulnerability of the dataset. Positive residuals indicate higher vulnerability, while negative residuals indicate lower vulnerability.

3 Results

3.1 MCMA urban growth projection to 2060 under a no restrictive scenario

SLEUTH simulations showed an expansion in urban area of nearly 4,790 km2 under the U2060 experiment. Urban expansion mainly occurred in the State of Mexico and Hidalgo located in the northern portions of the MCMA, with nearly 2,241 km2 and 2,158 km2 of urban and built-up land by 2060, respectively (Fig 2—right). In Mexico City, urban expansion took place in the southern portion of the city where the conservation areas are located (Fig 2—left). Overall, total losses of 63% and 23% occurred in dryland and irrigated cropland and pasture cover categories, respectively. Remaining losses of 14% were associated to shrubland, barren or sparsely vegetated, grassland, water bodies, evergreen needleleaf and broadleaf forests land cover categories (Fig 2, S2 Table).

3.2 Projected changes in near-surface temperature under a no restrictive urban growth scenario

Projected changes in near-surface temperature over the MCMA were closely linked to urban sprawl. Heat exposure calculated as differences in Tmean, Tmax and Tmin between the U2060 and CNTRL experiments are shown in Fig 3 and Table 2. Simulated increases in Tmean were greater in the State of Mexico, followed by Hidalgo and Mexico City with 0.7°C, 0.5°C, and 0.4°C respectively. Tmax changes ranged from 0.2 to 0.3°C, with greatest changes observed in the State of Mexico. Overall, changes in Tmin were greater than changes observed for Tmax and Tmean. Tmin increases up to 1.3°C were observed in the State of Mexico, 1.0°C in Hidalgo, and 0.9°C in Mexico City (Table 2).

Fig 3. Spatial distribution of urban-induced heat exposure.

This was calculated as changes in a) Tmean, b) Tmax, and c) Tmin (°C) between the U2060 and CNTRL experiments for each administrative unit within the MCMA. States and municipalities shapefiles: Terms of use:

Table 2. Differences in maximum, minimum and mean temperature between the U2060 and CNTRL experiments for each administrative unit in Mexico City, the State of Mexico and Hidalgo.

3.2.1 Mexico City.

Tmean differences between U2060 and CNTRL experiments ranged from 0.1°C to 1.0°C (Table 2). The maximum change in Tmean was recorded for Tláhuac (C13; Fig 1A). Tmax variation was between 0.2°C and 0.8°C, with greatest temperature increases projected for Cuajimalpa (Table 2), located in northeast portion of Mexico City (C5; Fig 1). Simulated Tmin varied between 0.1°C and 2.4°C with changes above 1.0°C observed for Tláhuac (C13), Xochimilco (C16), Venustiano Carranza (C15), Iztacalco (C8) and Iztapalapa (C9) (Table 2; Figs 1A and 3).

3.2.2 State of Mexico.

Tmean differences between U2060 and CNTRL experiments were between 0.1°C to 1.3°C with greatest Tmean increases in Tepotzotlán (S14), Texcoco (S15), Ecatepec (S5) and Zumpango (S20) located in the northern and eastern portions of the State of Mexico (Table 2; Fig 1A). Simulated Tmax differences were between 0.1°C to 0.6°C and Tmin between 0.1°C to 2.5°C (Table 2; Fig 3). Roughly, 75% of administrative units in the State of Mexico showed Tmean, Tmax, and Tmin exceedances of 0.3°C, 0.1°C and 0.1°C, respectively, compared to the simulated temperatures projected for Mexico City (Table 2).

3.2.3 Hidalgo.

Tmean differences between U2060 and CNTRL experiments were between 0.1°C to 1.1°C, whereas Tmax and Tmin differences ranged between 0.1°C to 0.3°C and 0.1°C to 2.6°C, respectively (Table 2). Greater changes in Tmean, Tmax and Tmin were observed in 66% of administrative units, with greatest increases in Cuenca de México Hidalguense (H5) and Llano de Tula (H3), both located in the southern border of Hidalgo (Figs 1a and 3).

3.3 Relative contribution of different land cover conversion on projected temperature changes

CNTRL and U2060 land cover conditions within the selected transect are shown in Fig 4. The land cover categories represented within the selected transect included dryland and irrigated cropland and pasture, evergreen needleleaf and broadleaf forest and, to a lesser extent, shrubland and water bodies. Under the U2060 experiment, nearly 7% of dryland and irrigated cropland and pasture areas were replaced by urban land and built-up areas (-99.0°W to -99.2°W and -98.8°W to -98.9°W). Within the selected transect, shrubland and water bodies were completely lost under U2060 conditions. Changes in simulated Tmean, Tmax and Tmin differed according to specific land cover conversion (Fig 4 upper panel). The loss of irrigated/dryland cropland and pasture bordering urban areas showed Tmean, Tmax and Tmin increases up to 1.9°C, 0.8°C and 4.2°C, respectively, whereas conversion of water bodies by urban land up to 1.9°C, 1.8°C and 2.4°C, respectively (Fig 4; S2 Table).

Fig 4. Temperature behavior within the west-east transect at 19.8 °N.

Land cover conditions (lower panel) within the west-east transect at 19.8 °N as defined in Fig 2 under the CNTRL and UN2060 experiments. Different colors represent land cover categories according to the USGS scheme following Fig 2 palette. Upper panel shows the simulated Tmean (black), Tmax (red) and Tmin (blue) changes between CNTRL and U2060 associated to the conversion of different land cover categories by urban land and build-up areas. CNTRL and U2060 conditions are represented by full and dotted lines, respectively.

3.4 Socioeconomic vulnerability for identified administrative units greater exposed to increases in near-surface temperature

Projected Tmean was expected to increase >1.0°C and <1.5°C in nearly 17% (n = 7) of the administrative units evaluated (Table 2), specifically in Tláhuac (C13-Mexico City), Ecatepec, Zumpango, Tepozotlán, Texcoco (S5, S20, S14 and S15, respectively—State of Mexico), Cuenca de México Hidalguense and Llano de Tula (H5 and H3, respectively—Hidalgo). Overall, the Gower´s residuals showed that 57% (n = 4) of the administrative units greater exposed to temperature increases tended to present positive values for the population, economic, health and infrastructure dimensions of vulnerability. Exploratory analyses showed weak positive and negative correlations between projected temperatures and socioeconomic indicators (S1 Table). However, associations were only significant between simulated Tmax and the number of households without electricity (h_noelec; r = -0.33, p< 0.05) and without access to the sewage system (h_nosewage; r = -0.3, p<0.05) (S1 Table; S4 Fig).

Regarding the population, health and economic dimensions of socioeconomic vulnerability, Ecatepec, and Cuenca de Mexico Hidalguense had the highest positive residuals values for the indicators pop, pop60>, pop_nomed and pop_nowork (Fig 5) indicating greatest unemployed population densities over the age of 60 without access to public health services. Regarding the infrastructure dimension of socioeconomic vulnerability, Tláhuac in Mexico City had positive residual values for h_soil and h_noelec, indicating a greatest number of households on soil floors and without electricity, whereas Texcoco in the State of Mexico, had the highest residual value for h_nwater, indicating the greatest number of households without access to potable water. For h_nosewage, all identified administrative units had negative residuals, meaning that overall, households had access to the sewage system (Fig 5).

Fig 5. Gower’s residuals.

Estimated residuals for the set of socioeconomic indicators in the seven administrative units that are more exposed to increases in average temperature (Tmean), as identified in Table 2.

4 Discussion

In this work, we simulated future urban growth in Mexico City Metropolitan Area projected to 2060 under a no restrictive scenario (U2060 experiment) using the cellular automata model SLEUTH. We integrated SLEUTH outputs to WRF to evaluate heat exposure measured as Tmean, Tmax and Tmin changes linked to future urban conditions. Socioeconomic vulnerability was assessed for selected administrative units according to a set of population, health, economic and infrastructure indicators. The Gower’s residuals provided insights into the relative vulnerability levels of different administrative units based on selected indicators of population, health, economy, and infrastructure. Hence, our results contribute to the overall understanding of the physical connections between temperature increases and the expansion of impervious surfaces in urban spaces and provides information about the spatial distribution of heat exposure in MCMA. This characterization allowed us to prioritize the administrative units greater exposed projected heat risks within the MCMA.

According to SLEUTH simulations, urban growth in the MCMA is expected to cover an area of 6,811 km2 by 2060 under our no restrictive scenario. This represents a percentage change of 337% (~4,790 km2) compared to current conditions. According to simulations, the greatest urban expansion would occur in the northern and southern portions of the State of Mexico and Hidalgo, respectively. In Mexico City, most of the projected urban expansion would take place in the southern municipalities of Tláhuac and Xochimilco. Our results agree with historical urban growth patterns in the MCMA. According to [73], the MCMA experienced an accelerated urban and industrial expansion during the last decade of the 20th century, with substantial increases in population, infrastructure, and transportation. From 1995 to 2010, urban expansion took place predominantly in Ecatepec, Cuautitlán Izcalli, Ixtapaluca and Zumpango located in the north of the State of Mexico as well as in Tizayuca in Hidalgo, reaching the highest growth rates by 2015 (S2 Fig). During that period, new informal urban areas were also developed in the southern portion of Mexico City, where much of the conservation land (CL) is located. The CL covers nearly 50% of the total area of Mexico City and represents a space of high ecological value protected under federal laws and regulations [74].

As urban borders expand to meet population growth demands, the replacement of natural land surfaces by impervious surfaces and the loss of green spaces increase the regional heat storage capacity [4, 73]. Our WRF simulations indicate significant contributions of land use and land cover changes associated to new urban developments to increases in Tmean up to 1.3°C in many parts of the MCMA by 2060, particularly in the north of the State of Mexico and the south of Hidalgo. These regions are characterized by a rather flat topography compared to the more complex orography encompassing the western and eastern portions of the MCMA [75]. Disparities in topography can significantly contribute to the simulated increases in near-surface temperature extending well beyond the MCMA limits (Fig 3).

Urban-induced temperature increases have been largely reported in the scholarly literature. For instance, [36] documented mean temperature differences between 1990 and 2009 urban landscapes from 0.5 to 1°C. Additionally, [37, 76] carried out numerical simulations of thermal variations due to past land use changes in MCMA with temperature increases of up to 4°C. In our study, the most relevant urban-related impact was observed for nightime temperatures (Tmin) with an increase up to 2.6°C reflecting the nocturnal nature of heat release. This work adds to the growing number of regional WRF modelling studies examining the relative contribution of different land cover types on simulated temperature changes (e.g., the replacement of water bodies by urban cover) [4, 7, 76]. Greatest Tmin changes were associated with losses of agricultural land and water bodies as shown in Fig 4. In this line of thinking, [37, 58, 76] established a direct physical link between the observed historical warming and the drastic reduction of the lacustrine system that once covered much of the total surface area of the Basin of Mexico.

Our results show that vulnerability to projected heat exposure is unequally distributed within MCMA. In Ecatepec and Cuenca de México Hidalguense, the major socioeconomic vulnerability sources were defined by population, economic and health conditions, whereas in Tláhuac and Texcoco by infrastructure conditions (Fig 1A, Table 2). According to current projections, the State of Mexico and Hidalgo are anticipated to undergo a population growth of 12% and 8%, respectively by 2050 accompanied by significant increases in social deprivation ([77]; S2 Fig). Notably, the State of Mexico is expected to become the densely populated state at the national level by 2050. Population growth for Mexico City is projected to decrease, unlike the State of Mexico and Hidalgo, most likely due to lower birth rates and increased mobility to other states [77].

Our findings emphasize the multidimensional nature of vulnerability as stated by [15]. This matches with some environmental studies conducted in urban centers worldwide (i.e., [14, 78]) which argue that socially and economically disadvantage people are disproportionally affected by extreme heat exposure. Our findings also stress the importance of developing joint solutions so that extreme heat mitigation policies and social vulnerability reduction and adaptation policies (i.e., human, social, financial, physical, and technical) can reinforce each other. Future vulnerability assessments for sustainable urban planning need also to explore how heat risk exposure is negotiated in decision-making processes under contexts of uncertainty. These analyses could shed light into the complexity of interactions of coupled human-environmental systems and how multiple risks cascade spatially and temporally [79]. This paper provides spatially explicit information on differentiated urban-related projected heat exposures that can be incorporated in future research aiming at examining uncertainties associated with socio-demographic projections, governance and institutional elements (i.e. accountability, transparency, conflict resolution, participation, and inclusion), which can also drive socioeconomic vulnerability.

We encountered a few challenges during the implementation of SLEUTH. Firstly, due to the complex nature of urban growth patterns within the MCMA, the calibration of SLEUTH with historical data was challenging. Secondly, within the MCMA, administrative units are subject to local, state, or federal conservation regulations crafted to safeguard priority conservation areas from urban growth. Nonetheless, the effectiveness of these regulations varies significantly throughout the MCMA due to deficiencies in governance and lax enforcement of land tenure and land regularization [80]. As a result, SLEUTH was not employed as a predicting model but was utilized primarily as an exploratory tool, facilitating the examination of the most extreme scenario—where urban sprawl encountered no constraints—Thirdly, the spatial resolution of our analysis proved challenging, particularly for the percentage slope allowed. Some areas of the MCMA have shown urban growth in slopes of 30% or more, measured at a scale not available for all input data layers. Thus, working at a 1 ha resolution might diminish SLEUTH capacity to capture the actual growth on some of the high-risk urban populations. Finally, in recent years some areas of the MCMA are growing vertically rather than through urban sprawl and the effects on extreme heat exposure cannot be measured with the two models applied.

Improvements to our methodology can be disaggregated into the two models used. Initially, the recent availability of openly accessible imagery sources, such as Sentinel and the new digital elevation models at a 1: 20,000 scale could have offered the potential to enhance the data resolution. However, this could have limited SLEUTH forecasting capacity by reducing the time frame of historical data. For WRF, improvements to our methodology include tuning the numerical weather prediction (NWP) based on the WRF model for the Valley of Mexico to capture model sensitivities to variations in how land surface processes associated to different land cover types are represented and parametrized (i.e., [36]). Additionally, important sources of variation in urban temperatures including anthropogenic heat release, urban density and design could be also included in future studies. We acknowledge that urban growth combined with climate change can lead to higher projected increases in near-surface temperature [4]. Analyzing synergistic impacts of multiple drivers of heat risk exposure is therefore critical to achieve sustainable urban systems, however, it was out of the scope of this paper.

5 Conclusions

Risk informed decision-making for sustainable urban planning requires the consideration of scenarios of exposure and vulnerability. In this work, we coupled WRF with SLEUTH to ascertain the vulnerability of urban populations to future heat exposure in Mexico City Metropolitan Area under a no restrictive urban growth scenario. According to SLEUTH simulations, greatest urban expansion occurred in the northwest part of the State of Mexico and the southern portion of Hidalgo. WRF simulations showed that overall, losses of natural surfaces increased near-surface temperature (Tmean, Tmax, and Tmin). Tmin was a better indicator of thermal changes compared to Tmax, Tmean. Urban-induced heat exposure together with socioeconomic attributes including population, health, economic status, and infrastructure are drivers of population vulnerability. In the case of Ecatepec and Cuenca de Mexico Hidalguense, socioeconomic vulnerability was mostly associated with population density, unemployed population, and population with limited access to health care, whereas Tláhuac vulnerability was associated with limited access to infrastructure services such as drinking water and electricity. The results of this study can serve as reference information for decision-makers to ensure that urban expansion in the MCMA is sustainable through the implementation of urban policy measures that regulate the development of new urban settlements. This study emphasizes the role of urban planners in designing projects considering a wider extension beyond the MCMA to take preventive measures to reduce the environmental impacts and the consequences of warming. Moreover, close collaboration between academia and the government through long-term, well-financed research is required to better inform guidelines to mitigate heat exposure risks and reduced associated vulnerabilities.

Supporting information

S2 Fig. Historical and projected population growth from 1970 to 2030 for Mexico City (o), the State of Mexico (+) and Hidalgo (*).


S4 Fig.

Relationship between simulated Tmax and socioeconomic indicators: no. households without access to electric and sewage services (left and right panels respectively). Correlation coefficient estimates and associated p values are also shown.


S1 Table. Pearson correlation estimates (r) calculated between simulated temperatures (Tmean, Tmax, Tmin) and socioeconomic indicators: Total population (pop), population older than 60 years of age (pop60>), unemployed population (pop_nowork), population without access to health services (pop_nomed), number of households with soil floors (h_soil), without electricity (h_noelec), without access to drinking water (h_nwater), and without drainage (h_nosewage).


S2 Table. Quantitative impacts considering transformations of categories of vegetation and water to urban land.



We acknowledge the support of Dr. Hallie Eakin of the School of Sustainability, Arizona State University (ASU) during the development of this project. We would also like to thank Fidel Serrano-Candela, Rodrigo García Herrera, and Edith P. Villa Mendoza from the Laboratorio Nacional de Ciencias de la Sostenibilidad for their key participation in generating the urban projections, and to Octavio Gómez Ramos from the Instituto de Geofísica for his technical assistance with the WRF model. The authors gratefully acknowledge the computing time granted by LANCAD and CONACYT on the supercomputer Miztli at DGTIC UNAM (Project: LANCAD-UNAM-DGTIC-393).


  1. 1. Oke T. Boundary Layer Climates. Second edi. Journal of Chemical Information and Modeling. Taylor$Francis; 1987.
  2. 2. Seto KC, Reenberg A, Boone CG, Fragkias M, Haase D, Langanke T, et al. Urban land teleconnections and sustainability. Proc Natl Acad Sci. 2012;109: 7687–7692. pmid:22550174
  3. 3. Seto KC, Güneralp B, Hutyra LR. Global forecasts of urban expansion to 2030 and direct impacts on biodiversity and carbon pools. Proc Natl Acad Sci U S A. 2012;109: 16083–16088. pmid:22988086
  4. 4. Chapman S, Watson JEM, Salazar A, Thatcher M, McAlpine CA. The impact of urbanization and climate change on urban temperatures: a systematic review. Landsc Ecol. 2017;32: 1921–1935.
  5. 5. Shem W, Shepherd M. On the impact of urbanization on summertime thunderstorms in Atlanta: Two numerical model case studies. Atmos Res. 2009;92: 172–189.
  6. 6. Seto KC, Parnell S, Elmqvist T. A global outlook on urbanization. Urbanization, biodiversity and ecosystem services: Challenges and opportunities. Springer, Dordrecht; 2013. pp. 1–12.
  7. 7. Vinayak B, Lee HS, Gedam S, Latha R. Impacts of future urbanization on urban microclimate and thermal comfort over the Mumbai metropolitan region, India. Sustain Cities Soc. 2022;79: 103703.
  8. 8. Kalnay E, Cai M. Impact of urbanization and land-use. Nature. 2003;425: 528–531.
  9. 9. Yu Z, Yao Y, Yang G, Wang X, Vejre H. Strong contribution of rapid urbanization and urban agglomeration development to regional thermal environment dynamics and evolution. For Ecol Manage. 2019;446: 214–225.
  10. 10. Adachi SA, Kimura F, Kusaka H, Duda MG, Yamagata Y, Seya H, et al. Moderation of summertime heat island phenomena via modification of the Urban form in the Tokyo metropolitan area. J Appl Meteorol Climatol. 2014;53: 1886–1900.
  11. 11. Chen F, Yang X, Zhu W. WRF simulations of urban heat island under hot-weather synoptic conditions: The case study of Hangzhou City, China. Atmos Res. 2014;138: 364–377.
  12. 12. Heaviside C. Urban Heat Islands and their Associated Impacts on Health. Oxford University Press; 2020.
  13. 13. Hans-O. Pörtner, Debra C. Roberts H. Climate Change 2022—Impacts, Adaptation and Vulnerability—Summary for Policymakers. Ipcc. 2022.
  14. 14. Tuholske C, Caylor K, Funk C, Verdin A, Sweeney S, Grace K, et al. Global urban population exposure to extreme heat.
  15. 15. Adger WN. Vulnerability. Glob Environ Chang. 2006;16: 268–281.
  16. 16. Adger WN, Eakin Hallie, Winkels A. Nested and teleconnected vulnerabilities to environmental change. Frontiers in Ecology and the Environment. 2009. pp. 150–157. pmid:32313513
  17. 17. Mitchell BC, Chakraborty J. Landscapes of thermal inequity: disproportionate exposure to urban heat in the three largest US cities. Environ Res Lett. 2015;10: 115005.
  18. 18. Chakraborty T, Hsu A, Manya D, Sheriff G. Disproportionately higher exposure to urban heat in lower-income neighborhoods: A multi-city perspective. Environmental Research Letters. Institute of Physics Publishing; 2019.
  19. 19. García-Burgos J, Miquelajauregui Y, Vega E, Namdeo A, Ruíz-Olivares A, Mejía-Arangure JM, et al. Exploring the Spatial Distribution of Air Pollution and Its Association with Socioeconomic Status Indicators in Mexico City. Sustain. 2022;14.
  20. 20. Turner BL, Kasperson RE, Matsone PA, McCarthy JJ, Corell RW, Christensene L, et al. A framework for vulnerability analysis in sustainability science. Proc Natl Acad Sci U S A. 2003;100: 8074–8079. pmid:12792023
  21. 21. Eakin H, Janssen MA, Anderies JM. Robustness Trade-offs in Social-Ecological Systems. 2007;1: 43–65.
  22. 22. Tellman B, Bausch JC, Eakin H, Anderies JM, Mazari-Hiriart M, Manuel-Navarrete D, et al. Adaptive pathways and coupled infrastructure: Seven centuries of adaptation to water risk and the production of vulnerability in Mexico city. Ecol Soc. 2018;23.
  23. 23. IPCC. AR6 Synthesis Report. Diriba Korecha Dadi. Panmao Zhai; 2023.
  24. 24. Schönhart M, Trautvetter H, Parajka J, Blaschke AP, Hepp G, Kirchner M, et al. Modelled impacts of policies and climate change on land use and water quality in Austria. Land use policy. 2018;76: 500–514.
  25. 25. Verburg PH, Eickhout B, Van Meijl H. A multi-scale, multi-model approach for analyzing the future dynamics of European land use. Ann Reg Sci. 2008;42: 57–77.
  26. 26. Das kleinskalige Strömungsmodell MUKLIMO_3 Teil 2: Thermodynamische Erweiterungen.
  27. 27. De Ridder K, Lauwaet D, Maiheu B. UrbClim—A fast urban boundary layer climate model. Urban Clim. 2015;12: 21–48.
  28. 28. Lac C, Chaboureau JP, Masson V, Pinty JP, Tulet P, Escobar J, et al. Overview of the Meso-NH model version 5.4 and its applications. Geosci Model Dev. 2018;11: 1929–1969.
  29. 29. Lauwaet D, Hooyberghs H, Maiheu B, Lefebvre W, Driesen G, Van Looy S, et al. Detailed urban heat island projections for cities worldwide: Dynamical downscaling CMIP5 global climate models. Climate. 2015;3: 391–415.
  30. 30. Oswald SM, Hollosi B, Žuvela-Aloise M, See L, Guggenberger S, Hafner W, et al. Using urban climate modelling and improved land use classifications to support climate change adaptation in urban environments: A case study for the city of Klagenfurt, Austria. Urban Clim. 2020;31.
  31. 31. Kong J, Zhao Y, Carmeliet J, Lei C. Urban heat island and its interaction with heatwaves: A review of studies on mesoscale. Sustain. 2021;13.
  32. 32. Fallmann J, Suppan P, Emeis S. Modeling of the Urban Heat Island (UHI) using WRF-Assessment of adaptation and mitigation strategies for the city of Stuttgart. EGU General Assembly Conference Abstracts. 2013. pp. EGU2013–4251.
  33. 33. Mohan M, Bhati S. Analysis of WRF Model Performance over Subtropical Region of Delhi, India. Adv Meteorol. 2011;2011: 1–13.
  34. 34. Chen F, Kusaka H, Bornstein R, Ching J, Grimmond CSB, Grossman-Clarke S, et al. The integrated WRF/urban modelling system: Development, evaluation, and applications to urban environmental problems. Int J Climatol. 2011;31: 273–288.
  35. 35. Clarke KC, Hoppen S, Gaydos L. A self-modifying cellular automaton model of historical urbanization in the San Francisco Bay area. Environ Plan B Plan Des. 1997;24: 247–261.
  36. 36. López-Espinoza ED, Zavala-Hidalgo J, Gómez-Ramos O. Weather forecast sensitivity to changes in urban land covers using the WRF model for central México. Atmósfera. 2012.
  37. 37. Benson-Lira V, Georgescu M, Kaplan S, Vivoni ER. Loss of a lake system in a megacity: The impact of urban expansion on seasonal meteorology in Mexico City. J Geophys Res Atmos. 2016;121: 3079–3099.
  38. 38. López-Bravo C, Caetano E, Magaña V. Forecasting summertime surface temperature and precipitation in the Mexico City metropolitan area: Sensitivity of the WRF model to land cover changes. Front Earth Sci. 2018;6: 6.
  39. 39. Aquino-Martínez LP, Quintanar AI, Ochoa-Moya CA, López-Espinoza ED, Adams DK, Jazcilevich-Diamant A. Urban-Induced Changes on Local Circulation in Complex Terrain: Central Mexico Basin. Atmosphere (Basel). 2021;12: 904.
  40. 40. Nations U. The World’s cities in 2018. Dep Econ Soc Aff Popul Div World Urban Prospect. 2018; 1–34.
  41. 41. Graizbord B, Granillo JLG. Urban Growth and Environmental Concerns: The Venture of the Greater Mexico City Metropolitan Area. Polit Policy. 2019;47: 178–206.
  42. 42. Molina LT, de Foy B, Mart\’\inez OV, Paramo Figueroa VH. Air quality, weather and climate in Mexico City. World Meteorol Organ Bull. 2009;58: 48.
  43. 43. Jáuregui E. The heat spells of Mexico City. Investig Geogr. 2009;70: 71–76.
  44. 44. López Vázquez VH, Plata Rocha W. Análisis de los cambios de cobertura de suelo derivados de la expansión urbana de la Zona Metropolitana de la Ciudad de México, 1990–2000. Investig Geogr. 2009;68: 85–101.
  45. 45. Di Pietro S. Urban Shrinkage and Suburbanization in Mexico: A View Based on the Last Census Data. Iberoam J Lat Am Caribb Stud. 2021;50.
  46. 46. Ziccadi A. Poverty and urban inequality: the case of Mexico City metropolitan region. Int Soc Sci J. 2016;217: 205–219.
  47. 47. Jantz CA, Goetz SJ, Shelley MA. Using the SLEUTH urban growth model to simulate the impacts of future policy scenarios on urban land use in the Baltimore-Washington metropolitan area. Environ Plan B Plan Des. 2004;31: 251–271.
  48. 48. Mahiny AS, Clarke KC. Guiding SLEUTH land-use/land-cover change modeling using multicriteria evaluation: Towards dynamic sustainable land-use planning. Environ Plan B Plan Des. 2012;39: 925–944.
  49. 49. Clarke KC. A Decade of Cellular Urban Modeling with SLEUTH: Unresolved Issues and Problems. Plan Support Syst Cities Reg. 2008; 47–60.
  50. 50. Bajracharya P, Lippitt CD, Sultana S. Modeling Urban Growth and Land Cover Change in Albuquerque Using SLEUTH. Prof Geogr. 2016;72: 181–193.
  51. 51. USGS, UCSB. Project Gigalopolis.
  52. 52. INEGI (Instituto Nacional de Estadística y Geografía). Conjunto de elevaciones mexicano CEM 3.0. 2013.
  53. 53. Instituto Nacional de Estadística y Geografía (INEGI). Uso del suelo y vegetación, escala 1:250,000, Serie I (continuo nacional). 1980.
  54. 54. Instituto Nacional de Estadística y Geografía (INEGI). Uso del suelo y vegetación, escala 1:250000, Serie III (continuo nacional). 2000.
  55. 55. Instituto Nacional de Estadística y Geografía (INEGI). Uso del suelo y vegetación, escala 1:250,000, Serie VI (continuo nacional). 2015.
  56. 56. INEGI (Instituto Nacional de Estadística y Geografía). Red Nacional de Caminos y conjunto de datos vectoriales de carreteras y vialidades urbanas Distrito Federal, Estado de México, Morelos, Hidalgo, Puebla y Tlaxcala. 2018.
  57. 57. Dietzel C, Clarke KC. Toward optimal calibration of the SLEUTH land use change model. Trans GIS. 2007;11: 29–45.
  58. 58. López-Espinoza ED, Zavala-Hidalgo J, Mahmood R, Gómez-Ramos O. Assessing the impact of land use and land cover data representation on weather forecast quality: A case study in central Mexico. Atmosphere (Basel). 2020;11: 1242.
  59. 59. Saha S, Moorthi S, Wu X, Wang J, Nadiga S, Tripp P, et al. The NCEP climate forecast system version 2. J Clim. 2014;27: 2185–2208.
  60. 60. Stauffer DR, Seaman NL, Binkowski FS. Use of four-dimensional data assimilation in a limited-area mesoscale model Part II: effects of data assimilation within the planetary boundary layer. Mon Weather Rev. 1991;119: 734–754.
  61. 61. Deng A, Stauffer DR, Zielonka JR, Hunter GK. 2.5 A COMPARISON OF HIGH-RESOLUTION MESOSCALE FORECASTS USING MM5 AND WRF-ARW. 22nd Conf Weather Anal Forecast Conf Numer Weather Predict. 2007.
  62. 62. Kain JS. The Kain–Fritsch convective parameterization: an update. J Appl Meteorol. 2004;43: 170–181.
  63. 63. Skamarock WC, Klemp JB, Dudhia J, Gill DO, Liu Z, Berner J, et al. A description of the advanced research WRF model version 4. Natl Cent Atmos Res Boulder, CO, USA. 2019; 145.
  64. 64. Chen F. The Noah Land Surface Model in WRF: A short tutorial. 2007.]
  65. 65. Lai D, Lian Z, Liu W, Guo C, Liu W, Liu K, et al. A comprehensive review of thermal comfort studies in urban open spaces. Sci Total Environ. 2020;742: 140092. pmid:32640397
  66. 66. Loveland TR, Reed BC, Brown JF, Ohlen DO, Zhu Z, Yang L, et al. Development of a global land cover characteristics database and IGBP DISCover from 1 km AVHRR data. Int J Remote Sens. 2000;21: 1303–1330.
  67. 67. Shi L, Moser S. Transformative climate adaptation in the United States: Trends and prospects. Science (80-). 2021;372: eabc8054.
  68. 68. Instituto Nacional de Estadística y Geografía (INEGI). Censo de Población y Vivienda. 2020.
  69. 69. Gower JC. Some distance properties of latent root and vector methods used in multivariate analysis. Biometrika. 1966;53: 325–338.
  70. 70. Digby PGN, Kempton RA. Ecological data. Multivariate analysis of ecological communities. Springer; 1987. pp. 1–26.
  71. 71. Bojórquez-Tapia LA, Ongay-Delhumeau E, Ezcurra E. Multivariate approach for suitability assessment and environmental conflict resolution. J Environ Manage. 1994;41: 187–198.
  72. 72. Masson-Delmotte V, P. Z, H.O. Pörtner D, Roberts J. Skea PR, Shukla A, et al. Informe especial del IPCC: Calentamiento global de 1,5 °C. Resum para Responsab políticas Organ Meteorológica Mund. 2018; 1–32.
  73. 73. Romero NC. Modelo espacial y pronóstico de la mancha urbana. Geogr Urbana. 2016.
  74. 74. Aguilar AG, Flores MA, Lara LF. Peri-Urbanization and Land Use Fragmentation in Mexico City. Informality, Environmental Deterioration, and Ineffective Urban Policy. Front Sustain Cities. 2022;4: 1–19.
  75. 75. Jauregui E. Heat island development in Mexico City. Atmos Environ. 1997;31: 3821–3831.
  76. 76. Jazcilevich A, Fuentes V, Jauregui E, Luna E. Simulated urban climate response to historical land use modification in the Basin of Mexico. Clim Change. 2000;44: 515–536.
  77. 77. CONAPO. Proyecciones de la población de México y de la entidades federativas 2016–2050 (República Méxicana). Cons Nac Población. 2019; 66.
  78. 78. Yang S, Ding L, Prasad D. A multi-sector causal network of urban heat vulnerability coupling with mitigation. Build Environ. 2022;226.
  79. 79. Simpson NP, Mach KJ, Constable A, Hess J, Hogarth R, Howden M, et al. A framework for complex climate change risk assessment. 2021; 489–501.
  80. 80. Tellman B, Eakin H, Turner BL. Identifying, projecting, and evaluating informal urban expansion spatial patterns. J Land Use Sci. 2022;17: 100–112.
  81. 81. Ruiz Angulo A, López Espinoza ED. Estimación de la respuesta térmica de a cuenca lacustre del Valle de México en el siglo XVI: un experimento numérico. Boletín de la Sociedad Geológica Mexicana. 2015;67(2): 215–225.