The Groundwater Management in the Mexico Megacity Peri-Urban Interface

Abstract

Megacities boost peri-urban socioeconomic development but fulfill their high natural resource demands by overexploitation, yielding irreversible environmental damage in surroundings that turn into sacrifice zones. This study reports the effects on the Cuautitlán-Pachuca Valley, the Mexico City main expansion zone at the northeast of the metropolitan area on the Central Mexico plateau, the trend scenarios from 2020 to 2050, and the actions to mitigate the growing water demand that will worsen its aquifer overexploitation. We designed a conceptual archetype to apply the Water Evaluation and Planning System (W.E.A.P.) mathematical model calibrated with 2013–2014 data to calculate groundwater volume demand in future scenarios. The demand output for the international airport and agriculture was less than 5%. The local climate change effect up to 2050 will slightly reduce the infiltration. The most crucial water demand increase (195% in 2050) is due to the population and industrial growth of the Mexico City northern municipalities (89% of the total groundwater extraction volume), and the aquifer will have a notable −2192.3 hm³ accumulated deficit in 2050, while urban sprawl will decrease water infiltration by 2.3%. Mitigation scenarios such as rainwater harvesting may reduce the urban water supply only by 9%, and a leak cutback will do so by 24%, which is still insufficient to achieve sustainable water management in the future. These outcomes emphasize the need to consider other actions, such as importing water from near aquifers and treating wastewater reuse to meet the future water demand.

1. Introduction

The increase in the average world temperature and extreme meteorological events are becoming more frequent. The decrease in rainfall increases the areas affected by drought because of climate change, having a higher impact on the water cycle and requiring urgent adaptation actions to achieve sustainable water management that is less expensive [1].

The groundwater consumption of more than 1.7 billion people surpasses recharge, increasing water stress and causing economic, social, and environmental impairment [2]. Sustainable Development Goal # 6, “Clean Water and Sanitation”, points out the need for efficient water management to reduce people suffering from the drinking water shortage forecast in 2030 [3]. In contrast, there will be 35 megacities (605 million inhabitants in total) in 2024 and 41 in 2030, so the water demand will still increase [4].

As a megacity surpasses ten million inhabitants [5,6], its high population produces high quantities of pollutant and greenhouse gas emissions, huge wastewater volume, solid wastes, and ravage resources such as water, energy, and food from the nearby areas [7].

The proximity to cities promotes economic growth, markets, services, cultural events, and job opportunities in the peri-urban areas [8]. However, their harmful effects on their conversion on waste discharge sites and the overexploited resources produce irreversible environmental damage that can turn them into sacrifice zones [9]. Water access is critical in megacities that depend on groundwater because the growing urban land use decreases the aquifers’ recharging while demand grows [10].

As reported in Beijing, China, the megacity’s water service regional infrastructure network extends beyond the city’s limits and highly depends on imported resources from the peripheral area [11]. In India, land use changes have negatively affected water management within the surrounding areas of Gurgaon and Delhi [12,13].

The peri-urban landscape transformation was reported in Latin America through the rapid growth of four developing cities, Santiago de Chile, Panama City, Mexico City, and Rio de Janeiro, with different challenges [14]. In contrast, the Mezquital Valley, northwest of Mexico City, has shown strong development in the agricultural and industrial sectors, which is allowed by the megacity sewage supply [15].

There is a greater interest in studying environmental problems and the demand for natural resources management within megacities, but very few studies report their influence on the peri-urban areas. Appendix A (

Table A1. The literature reports on the effects of megacities on their surroundings.

Title	Citation	Field
Differences in ozone photochemical characteristics between the megacity Nanjing and its suburban surroundings, Yangtze River Delta, China	[49]	Air
Seismic hazard assessment in the megacity of Blida (Algeria) and its surrounding regions using a parametric-historic procedure	[50]	Air
Anthropogenic inputs from a coastal megacity are linked to greenhouse gas concentrations in the surrounding estuary	[51]	Air
Differences in ozone photochemical characteristics between the megacity Tianjin and its rural surroundings	[52]	Air
Assessment of ambient aerosol sources in two important Atlantic Rain Forest hotspots in the surroundings of a megacity	[53]	Air
Introduction to the special issue “In-depth study of air pollution sources and processes within Beijing and its surrounding region (APHH-Beijing).”	[54]	Air
Exploring the variation of black and brown carbon during COVID-19 lockdown in megacity Wuhan and its surrounding cities, China	[55]	Air
The impact of circulation patterns on regional transport pathways and air quality over Beijing and its surroundings	[56]	Air
Formation of secondary organic aerosol in the Paris pollution plume and its impact on surrounding regions	[57]	Air
The water–energy nexus of megacities extends beyond geographic boundaries: a case of Beijing	[11]	Water
From the core to the periphery: conflicts and cooperation over land and water in peri-urban Gurgaon, India	[12]	Water
Impact of continuous Jakarta megacity urban expansion on the formation of the Jakarta–Bandung conurbation over the rice farm regions	[7]	Water/Soil
The food–water quality nexus in peri-urban aquacultures downstream of Bangkok, Thailand	[58]	Water/Soil
Megacity wastewater poured into a nearby basin: looking for sustainable scenarios in a case study	[15]	Water/Soil
Analyzing the effects of different scenarios on the surrounding environment in a high-density city	[59]	Soil
Impacts of urban expansion on relatively smaller surrounding cities during heat waves	[60]	Soil

) compiles some literature reports about these effects in those peripheral regions.

The Northern Zone of Mexico City

General. The Cuautitlán-Pachuca Valley (Figure 1) surface is about 3870 km², at 2390 m above sea level on the Mexican plateau. It holds the highly urbanized northeastern metropolitan zone of Mexico City (Mexico State) and the southwest of the Hidalgo State, which has the largest non-urban area. The city’s northern municipalities are among Latin America’s most highly populated regions [16]. Then, the Valley is virtually divided into two zones, the most urbanized in the Mexico State and the least in the Hidalgo State, which is meant to be the peripheral area for Mexico City’s expansion with an ever-growing population [17,18].

Land use. Agriculture land use prevails in the Hidalgo State at N.E., while in the Mexico State, there are two small irrigation districts (Chiconautla and La Concepción), and it has the highest concentration of technology clusters for the automotive, food, chemical, and plastic industrial sectors in the whole Valley (12 in Cuautitlán Izcalli, 10 in Tultitlán, 10 in Ecatepec, and 3 in Cuautitlán) [19].

Furthermore, the new Felipe Ángeles International Airport (A.I.F.A. by its Spanish acronym) is in the Mexico State (Zumpango municipality) (Figure 1). Its first stage covered air cargo transport and commercial domestic passenger flights. By 2042, it will be at its total capacity of eighty million passengers yearly [20]. It would aggravate the water demand by increasing the floating population and expanding urbanization since the plan considers the construction of a suburban train connecting with the megacity, which will attract new housing developments and offer new jobs within the Cuautitlán-Pachuca Valley.

Additionally, a 343 ha project, P.L.A.T.A.H., is speedily evolving in the Hidalgo State for industrial, logistic, commercial, and service areas, located at the intersection of the Northern Arch (highway connecting the Pacific Ocean and the Gulf of Mexico) with the México-Pachuca highway. P.L.A.T.A.H. also concentrates on international trade goods and road cargo transportation services [21].

Groundwater. The aquifer represents 23.6% of the Mexican Valley surface [22]. It is a semi-confined, heterogeneous, and anisotropic type, belonging to the Neo-Volcanic Axis physiographic region in the sub-province “Lakes and Volcanoes of Anahuac”. Reports show three zones in the subsurface with different hydrogeological characteristics. The water supply for custom sectors and new developments counts on the water supply from the aquifer despite its high overexploitation [23].

The cited studies reported the Cuautitlán-Pachuca aquifer as one of the country’s most overexploited in Mexico [21,22,23], and aggravating demand factors and mitigation actions were reported for the Hidalgo State area. However, the urban and industrial sectors of the Mexico State municipalities mentioned above were not considered, and it is a must.

The present study aims to achieve the water demand scenarios for the Cuautitlán-Pachuca Valley as a case study altered by the megacity’s proximity and expansion. The methodology approach defines conceptual and mathematical models based on sub-basins for the surface and groundwater volume calculation. We will simulate the water demand volumes in realistic scenarios up to the year 2050 with projected actions in the Valley, managed as disturbances in the software W.E.A.P. (2023.0), previously calibrated with 2013–2014 data. First, we will consider inertial scenarios perturbed by the urban sprawl, population, industry, and agriculture growth rates to predict water demand. Secondly, we include the local prognosis of climate change and the international airport demand. Finally, we consider two possible adaptation actions (rainwater harvest and urban tap water leak reduction) to assess the mitigating potential for water demand. Both are supposed to be the most favorable scenario. Their differences from the inertial one will support the convenience of those actions in public policies and the design of programs that may contribute to the sustainable use of water threatened by the growing demand.

This work contributes to the knowledge of the megacity’s effects on the sustainable use of water in their periphery. The methodology approach based on models can be suitable for studying peri-urban areas threatened by the growing demand for urban land use. From the local point of view, the study complements earlier reports because we grab information from the most populated and industrial regions of Mexico City located on the valley surface. The study will evaluate the accomplishment of the proposed actions to mitigate the growing overexploitation of the Cuautitlán-Pachuca aquifer.

2. Materials and Methods

2.1. Conceptual Model: Boundaries and Description

The most essential anthropogenic and natural features are depicted in Figure 1. The water inflows, outflows of surface water, and groundwater parameters from the demand sectors of the Cuautitlán-Pachuca Valley are represented in the scheme shown in Figure 2, which was used for the next mathematical model.

• The baseline year 2013 was selected, as it is an average year according to C.O.N.A.G.U.A.’s rainfall data and McKee standardized index methodology [24], and because most data are available. For calibration, minor 2014 data were included.
• The area is considered a watershed and divided into four hydrographic sub-basins, three highly urbanized in the Mexico State (Tepotzotlán, Cuautitlán, Texcoco-Zumpango) and the greatest with more extensive agricultural land use in Tezontepec in the Hildago State. The boundaries were defined according to the topographic national data [25]. See Figure 2.
• Each sub-basin has a primary tributary river where the runoff flows mixed with local wastewater.
• Rainfall is the primary water source, producing the runoff and infiltration that recharges the aquifer.
• A groundwater inflow comes from the Apan aquifer into the Cuautitlán-Pachuca aquifer with a constant horizontal flow of 5.7 hm³/year (C.O.N.A.G.U.A., 2024) [26,27].
• The Texcoco aquifer does not contribute to another horizontal flow because of its reported huge water table reduction [26].
• The main water outflows are evapotranspiration and groundwater pumping from the aquifer to supply urban, service, and industry demands. The locally produced wastewater is discharged into the Great Drain Channel and flows out of the Valley. The Great Drain Channel and the West Interceptor Tunnel cross the area without providing water, but the first receives the local sewage and enables its outflow from the model.
• The aquifer supplies the Chiconautla irrigation district.
• The reference evapotranspiration is calculated by the Penman–Monteith method [28].
• There are leaks in the tap water distribution system. The leak volumes in urban areas of the Mexico State are reported as 40% in [29] and 53% in the Hidalgo State [30].
• The model only includes the population in the northern portion of the Texcoco- Zumpango sub-basin that the Cuautitlán-Pachuca aquifer supplies since the other part is provided by the Mexico City aquifer [31].
• The model does not consider illegal extractions since no official reports of their volumes exist.
• Each sub-basin has a different average annual rainfall and mean annual temperature. Figure S1 in the Supplementary Materials shows maps with the distribution values.

Figure 2 shows the surface hydrology. The Cuautitlán River is highlighted since it comes from the western mountain range (with its affluents, Hondo Tepotzotlán River and San Pedro River) and flows towards the Mezquital Valley through the Tajo de Nochistongo and the Tequisquiac tunnels. In the rainy season, part of the waters is poured toward the Zumpango Lake through the Santo Tomas Channel [32]. Before leaving, the river fed the La Concepción and Guadalupe reservoirs, whose waters are used for the Concepción and Chiconautla irrigation districts.

The Great Drain Channel passes through the study area without providing surface water. Still, it collects the wastewater from the region’s urban settlements and industries before inflowing into the Mezquital Valley. The Great Drain Channel is an essential early infrastructure for the Mexico City sewage surface waters and the West Interceptor Tunnel, which flows underground [33].

Las Avenidas River is not perennial and collects wastewater from Pachuca City and nearby populations in the Tezontepec sub-basin. It flows through the plain towards Tizayuca, supplies sewage to El Manantial Reservoir for irrigation, and ends in the Zumpango lagoon.

2.2. Mathematical Model and Sub-Basins Definition

The Water Evaluation and Planning System (W.E.A.P.) was used for the volume calculations. It is recommended for the integrated approach to water resources planning [34] and contains the basic equations for water balance adjusted to urban and agricultural systems. It is ideal for predicting scenarios with chosen perturbations, such as the future of each sector’s supplies, water conservation, climatic effects, etc. The W.E.A.P. module “Rainwater runoff” was used to simulate the water cycle of the study area.

The topographic map allowed us to delimit the four sub-basins employed for the data introduction strategy using the W.E.A.P. schematic model shown in Figure S1 of the Supporting Information [25,35,36]. The calculation was performed for each sub-basin, and the Valley was obtained using the sum of the results.

2.2.1. Data for Water Balance Calculation: Baseline Year 2013

Rainfall and temperature.

The study used the C.L.I.C.O.M. daily climate data of the National Meteorological Service (S.M.N. according to its Spanish acronym) through its website platform of C.I.C.E.S.E. [37] for 2013 to 2015. The average values were obtained from the Climatic Records per state through the S.M.N. webpage with data from twenty-four weather stations in the study area. The data were interpolated (Kriging) in ArcGIS 10.5, yielding raster images with monthly and yearly rainfall, temperature, and evapotranspiration data for each sub-basin. Figures S2 and S3 in the Supporting Information show the raster maps.

2.2.2. Environmental Variables and Statistical Analyses

Evapotranspiration potential (ETo). It was calculated using the Penman–Monteith standardized method [28], accepted and recommended by the World Meteorological Organization, and adopted by the F.A.O. (see Equation (S1) in the Supplementary Materials). Figure S4 shows the raster map obtained.

Runoff and infiltration coefficients. The specific runoff coefficients (Cr) were calculated using the methodology reported by Galindo-Castillo et al. [21] by Equation (S2), which encompasses soil texture and slope, including land use. The mean slope of each sub-basin was calculated using the digital elevation map, texture data, and land use. Equation (S3) subtracted infiltration coefficients (Ci) in the Supplementary Materials. The W.E.A.P. model compiles each sub-basin’s Ci and Cr weighted mean values (Table S1).

Soil use and vegetation percentages. Each land use percentage introduced in the W.E.A.P. model was obtained from the vector information in the I.N.E.G.I. webpage [36] and later processed with ArcGIS 10.5. The results are shown in Table S2.

Water demand. The supply values were obtained from the C.O.N.A.G.U.A.’s Water Rights Public Records database [31]. They were disaggregated by municipality and sector and later organized by sub-basin (Table S3).

2.3. Validation of the Mathematical Model

The model was validated with the surface and groundwater results. The calculated water surface volumes were correlated with the actual volumes from the flow rate data of the Huehuetoca hydrometric station, code 26056, in the Cuautitlán River, Mexico State. The Pearson criteria linearly correlated the calculated vs. actual values for each month of 2013 and 2014 [38]. Also, Equation (1) calculates the mean absolute percentage error (M.A.P.E.) to indicate model precision [39].

$$MAPE=\frac{1}{n}{\sum }_{t=1}^n\frac{\left|A_t-F_t\right|}{A_t}\times 100$$

(1)

where n is the number of predictions, A_t is the actual value, and F_t is the forecast value. The groundwater results were validated by calculating the M.A.P.E. as previously described, with three recharging values previously reported by [26,29]. See Table S4.

2.4. Steady-State and Transient Conditions Scenarios

Business-As-Usual (B.A.U.) Inertial scenario. This scenario is unrealistic since it only includes the annual population growth rate calculated using a geometric model represented by Equation (2) [40]. Population data were from [41,42].

$$r={\left(\frac{N^f}{N^i}\right)}^{\frac{1}{k}}-1$$

(2)

where r is the population growth rate under the geometric assumption, Nⁱ and N^f correspond to the initial and final populations, and k is the number of years elapsing between both population values. This calculation was made for each sub-basin, summing the population data of each municipality involved for each initial (2010) and final year (2015). The average annual growth rate for each sub-basin is shown in Table S3 in the Supplementary Materials.

Inertial Reference (I.R.) scenario. It is the most realistic inertial scenario. It includes the previous scenario (B.A.U.) plus the water demands perturbation of the industrial, urban sprawl, and service growth rates. It is assumed that the population and service growth rates are the same. The increase in urban sprawl produces a decline in agricultural areas. Therefore, we assume that both rates represent the same absolute value but with opposite signs. The agricultural to urban land use change was calculated in this study with the QGis 3.18 software and C.O.N.A.B.I.O. vector layers [36]. Table S3 shows each sector’s growth rates for the sub-basins in percentages. This inertial scenario will be a reference because it includes the appropriate settings to compare perturbed scenarios.

Climate change effect (C.C.) scenario. It includes the previous I.R. scenario plus the reported and calculated data associated with climate change in the future (rainfall, mean temperature, and evapotranspiration) for each sub-basin. The municipality data were taken from the National Atlas of Vulnerability to Climate Change [43]. The temperature and rainfall values expected for 2030 and 2050 were taken from the SSP3 RCP 7 scenario, reported in the Coupled Model Inter-comparison Project (C.M.I.P.) [44], as observed in Table S4 in the Supplementary Materials. The rainfall values changes ranged between −1.52% and 0.26%, and those of temperature between 1.05 °C and 1.88 °C, up to 2060.

Felipe Ángeles International Airport (A.I.F.A.) scenario. Based on the above scenario (C.C.), we included the 750 m³/day water supply during the construction of the new airport (A.I.F.A.) in 2020 [45]. Moreover, we considered the passenger increase projection within two stages as reported by the Secretary for Agrarian, Territorial and Urban Development [20] and the Mexico City International Airport historical water supply data reported by the Secretary for Communications and Transportation [46]. The estimated demand is shown in Table S5 in the Supplementary Materials.

Rain harvesting (R.H.). Adaptation scenario 1: Based on the A.I.F.A. scenario, we included water demand mitigation through rainwater harvesting in urban areas from 2025. This action allows the reduction of annual tap water supply by up to 30%, according to Kim et al. [47] and van Dijk et al. [48]. The gradual installation of this system in urban areas is proposed to start at 10% in 2030 and up to 30% in 2050.

Avoiding leaks (A.L.). Adaptation scenario 2. Based on the A.I.F.A. scenario, we add the perturbation of water demand mitigation by increasing physical efficiency by leaking elimination in the urban areas’ water distribution system. For the Tezontepec sub-basin in Hidalgo, a 2% yearly infrastructure improvement is proposed from the 2020 to 2030 period [30] and 0.3% afterward up to 2050. In the Mexico State sub-basins, the physical efficiency would increase by 0.6% annually from 2025 to reach a final improvement of 75% in 2050.

The best adaptation scenario considered in the study is the sum of R.H. and AL.

3. Results and Discussion

3.1. Model Validation

The linear Pearson coefficient for surface water was 0.98, as observed in Figure S5, and the M.A.P.E. resulted in 11.5%. Groundwater yielded a M.A.P.E. value of 3.8%, and the results indicate that the mathematical model is valid. Table S6 shows the data.

3.2. Baseline Hydric Balance

Figure 3 shows the water balance data for each sub-basin, where the Tezontepec presents the highest evapotranspiration, infiltration, and runoff values due to its most significant area.

Sub-basins have larger runoff values than infiltration because of the high urbanization, mainly in the Cuautitlán and Tepotzotlán sub-basins. The land use volumes for the surface and groundwater in each sub-basin and the total Valley are shown in

Table 1. Areas (km²), groundwater, and surface water volumes (hm³) for land use demands in sub-basins and the Valley (baseline year 2013).

Sub-Basin	Area	Surface Water		Groundwater
		Agric. &	ID *	Agric. &	Ind. #	Urban	Services
Tezontepec	2062.2	8.0		2.0	1.3	128.8	0.2
Tx. Zumpango	1109.6	32.0	16.2	43.3	9.1	878.6	13.0
Tepotzotlán	420.0	28.1	0.8	8.4	2.7	138.0	5.0
Cuautitlán	384.9	45.2		5.9	8.6	46.8	6.0
Total Valley	3976.7	113.3	17.0	59.6	21.7	1192.2	24.2
Total surface water				Total groundwater
130.3				1297.0

* I.D. Irrigation District. ^& Agricultural. ^# Industrial.

.

The results previously reported by Galindo et al. [22] presented the analysis of the Tezontepec sub-basin in the Hidalgo State and did not include the highly populated Mexico State areas. Table 1 shows that they are the most prevailing, considering their high groundwater demands. Also, the surface water volumes are significantly smaller than groundwater’s, demonstrating the need to focus on managing the aquifer.

Table 2. Water balance components of the Cuautitlán-Pachuca Valley (4426 km²) in the baseline year 2013. Calculated by the W.E.A.P. model (hm³).

Component	P	Eto	I	R	GD	SD
Volume	1021.1	190.9	373.4	456.8	1297.0	130.3

P: precipitation; Eto: evapotranspiration; I: infiltration; R: runoff; GD: groundwater demand; SD surface water demand.

presents the results for the entire Cuautitlán-Pachuca Valley.

3.3. Steady-State and Transient Conditions Scenarios

Firstly, the simulated water balance and demands for the inertial scenarios (B.A.U. and I.R.) are modeled up to the year 2050, and the increasing water supply sector components are calculated. The actions included in the transient scenarios are added as perturbances to the I.R., and the volume results are shown in

Table 3. Groundwater and surface water supply volumes for each scenario, sub-basin, and the whole Valley (hm³).

	Scenarios	2013	2030						2050
Sub-Basin		B.L.	B.A.U.	I.R.	C.C.	A.I.F.A.	R.H.	A.L.	B.A.U.	I.R.	C.C.	A.I.F.A.	R.H.	A.L.
Tezontepec	G.W.S.	132	569	569	569	569	557	388	1317	1318	1318	1318	1201	827
	S.W.S.	8	8	8	8	8	8	8	8	8	8	8	8	8
Texcoco-Zumpango	G.W.S.	942	2106	2105	2105	2106	2063	2009	2985	2990	2990	2994	2736	2409
	S.W.S.	48	61	95	95	95	95	95	61	142	142	142	142	142
Cuautitlán	G.W.S.	67	132	139	139	139	137	133	161	179	179	179	167	145
	S.W.S.	43	43	43	43	43	43	43	43	43	43	43	43	43
Tepotzotlán	G.W.S.	154	383	387	387	387	380	369	618	633	633	633	580	509
	S.W.S.	29	30	30	30	30	30	30	30	29	29	29	29	29
Valley	G.W.S.	1296	3189	3199	3199	3201	3136	2899	5081	5119	5119	5123	4684	3890
	S.W.S.	128	142	175	175	175	175	175	142	222	222	222	222	222

G.W.S. Groundwater supply. S.W.S. Surface water supply. B.L. Baseline. B.A.U. Business-as-usual. I.R. Reference. C.C. Climate change. A.I.F.A. Airport. R.H. Rainwater harvesting. A.L. Avoiding leaks.

. Figure 4 and

Table 4. Cumulative effects on the aquifer water volumes by different disturbances in 2030 and 2050 (hm³).

Disturbance	Effect	2030	2050	Net Effect in 2050 *
Population	Increases demand	−1893	−3785	−3864.3
Urban sprawl	Decreases infiltration	−22.1	−25.7
Industries/Services	Increase demand	−10	−38
Climate change	Decreases infiltration	−1.9	−11.6
Airport	Increases demand	−2	−4
Rainwater harvesting	Decreases the demand	65	439	1672
Repairing leaks		302	1233

* Net overexploitation surplus: −2192.3 hm³.

show the surplus differences of perturbations. Table S7 displays the changes in the water cycle parameters.

3.3.1. Steady-State Scenarios (Inertial)

B.A.U. The B.A.U. scenario for 2030 and 2050 shows increases in groundwater demand due to population growth from +1893 hm³ and +3785 hm³, respectively, since the 2013 baseline year (1296 hm³). The huge rise in population demands is the highest of all anthropogenic effects in this study. Nevertheless, the future population growth rate (2030–2050) might be smaller than the current value shown in Table S3, but no such predictions have been reported, and B.A.U. calculates the volume demand for the worst scenario.

Reference (I.R.). Compared to the B.A.U. scenario, the I.R. shows a new surplus in groundwater demand of +10 hm³ and +38 hm³ for 2030 and 2050, respectively, produced by industrial and service growth. The total demand (including B.A.U.’s population) increase yields of +1903 hm³ in 2030 and +3823 hm³ in 2050 will make the unsustainable aquifer extraction.

With the same climate data I.R. and B.A.U for 2030 and 2050, evapotranspiration decreases by −48.7 hm³ and −56.8 hm³ due to the change in agriculture-to-urban land use. Also, the runoff increased by +26.6 hm³ and +31.1 hm^3, while infiltration was reduced by −22.1 hm³ and −25.7 hm³.

This reference scenario is helpful for comparisons with the subsequent transient scenarios because it is more realistic than B.A.U. despite the magnified effect of the current growth rates employed in Table S3.

3.3.2. Transient Conditions Scenarios

The volume results for inertial and transient scenarios are summarized in Table 3 for 2013 (baseline), 2030, and 2050. Table 4 and Figure 4 focus on the perturbation’s effects, leading to further conclusions.

Climate change scenario (C.C.). This scenario introduces the perturbance of predicted temperature and rainfall changes over time. For 2030 and 2050, the calculated rainfall decline is −3.7 hm³ and −17.2 hm³, with an increase in evapotranspiration of +0.6 hm³ and +0.5 hm³, respectively. Consequently, the aquifer infiltration decreases by −1.9 hm³ and −8.0 hm³ for both years, which turns to a recharge deficit of −2.3% in 2050 compared to the I.R. scenario. Also, the runoff decreases by −2.4 hm³ and −9.8 hm³, which may affect the not-irrigated agriculture in the Tezontepec sub-basin. See Figure S6. The C.C. harm is almost negligible compared to the anthropogenic effect produced by the population and industry in the I.R. scenario.

Airport scenario (A.I.F.A.). The increase in groundwater demand for passenger services in 2030 and 2050 resulted in +2 hm³ and +4 hm³, respectively, compared to the I.R. scenario. This value and the infiltration deficit due to C.C. and the huge groundwater demand in the I.R. scenario slightly raise the aquifer overexploitation.

Adaptation scenario. Rainwater harvesting (R.H.). Rainwater harvesting is progressively introduced in the model by decreasing the groundwater extraction over time, as explained in Section 2.4. This action reduces the demand by −63 hm³ and −439 hm³ in 2030 and 2050, respectively, compared to the A.I.F.A. scenario, and it is not enough to compensate for the growing groundwater demand. According to the rainwater-use calculator, it would decrease only by 9% in 2050 for the urban population (C.A.P.S.U.S.). The Valley is not a rainy zone, as it has about 7–8 months of a dry season with occasional rain in winter. More actions are needed to mitigate or restore the overexploited aquifer since the infiltration remains lower than the demand, and the following adaptation action is evaluated.

Adaptation scenario. Avoiding Leaks (A.L.). The gradual repair of leaks in the urban areas is explained in Section 2.4. This action has environmental and economic importance, and the results show a more significant effect than the R.H. scenario since the demand is reduced by −302 hm³ and −1233 hm³ for 2030 and 2050, respectively, compared to the A.I.F.A. scenario. Still, it is not enough to compensate for the groundwater demand, and even the sum of R.H. and A.L. would not do either (see Table 4 and Figure 4), as the sum of both adaptation scenarios would reduce the demand for the A.I.F.A. scenario by 32.6% in 2050. Sustainable aquifer management is not achieved, and the proposed actions mitigate but do not avoid the aquifer overexploitation. The infiltration needs to be improved, and stakeholders and policymakers should evaluate groundwater importation.

3.4. Scenarios Comparison

Figure S7 shows the flow water balance in Sankey diagrams, calculated for various scenarios, allowing for a visualization of the most relevant changes in 2050. They are included in the Supplementary Materials and give visual progress of the balance impairment over time.

Figure 4 illustrates the insufficiency of the groundwater demand mitigation actions compared with the baseline, which has been in an overexploitation condition since 2004, as the first studies reported [23].

Table 4 displays the effect of each scenario perturbation for 2030 and 2050, illustrating their hierarchy and the unsuccessful net effect. Those results demonstrate that population growth is the greatest threat to the aquifer, followed by the industrial, services, and commercial sectors. The sectorized analysis helps hierarchize the actions in political programs to achieve sustainability.

Surprisingly, the increasing demands of the industry/services sector are not critical, with some reliability uncertainties about the official data, of which sources based on private reports would be somehow incomplete or biased. The population is the most significant target for mitigating water demand.

The effects of climate change and airport presence are the least harmful. The adaptation actions proposed in this study are not enough to avoid the irreversible deterioration of the aquifer.

3.5. Limitations and Future Perspectives

The study was performed using a modeling approach based on available official data. The base year industrial water supply data were limited; hence, the real water demand would be higher than the calculated one, which is this study’s main limitation. Predictions are not exact in any modeling study, but in this case, considering the rising overexploitation of the aquifer, its damage will be for sure, and considerable water scarcity will happen.

The outlook of the water availability in the valley is upsetting but solvable if actions are taken in time, such as infiltration wells, water reuse, environmental education, and groundwater importation. They should be urgently planned with restricted legal protection of the aquifer.

4. Conclusions

Surface water does not play a relevant role in the Cuautitlán-Pachuca Valley, and its primary use is in agriculture. In contrast, groundwater is the foremost water source, but its unsustainable management is due to continuous aquifer overexploitation from the past, and this trend will jeopardize future water availability. It is the main problem detected in the area. The anthropogenic effects are the most significant threats, particularly from the utmost population growth in the northern zone of Mexico City, which demands around 89% of total groundwater extraction. In 2050, the valley population would increase the water demand by an outstanding 195%, and the urban sprawl growth would decrease the water infiltration by 2.3%. Consequently, the aquifer would have a −2192.3 hm³ deficit in 2050.

Climate change would little modify the water cycle of the Cuautitlán-Pachuca Valley, particularly that of Tepotzotlán in the Hidalgo State and the Mexico State sub-basins, and the international airport will slightly raise the groundwater demand. Both perturbations were not relevant for the future. The proposed mitigation actions do not solve the unsustainable management of the aquifer. Rainwater harvesting may reduce the urban water supply only by 9%, and a leak cutback will do so by 24% in 2050. Despite contributing to the conservation of aquifers, they are not close enough to compensate for the growing population’s demand. The surplus accumulated groundwater extraction was estimated to be more than 2192 hm³ in 2050 after mitigation by rain harvest and leak reparation. The huge overexploitation would potentially jeopardize the aquifer before 2050, leading to its irreversible damage, the area’s main problem. Thus, other relevant adaptation actions must be implemented, such as importing from other aquifers and reusing wastewater after reliable treatment. The modeling approach employed may be helpful for other studies in peri-urban areas.