Global basic Water-Service Access Index (GWSAI)

This index allows evaluating the impact of the coverage, quality and efficiency of drinking water and sanitation services, in terms of access to drinking water and sanitation services, and is evaluated by the following indicators: access to drinking water services, calculated with the information referring to: urban drinking water coverage (%), rural drinking water coverage (%) and disinfected water (% ). The indicator of access to sanitation services is composed of urban sewerage coverage (%), rural sewerage coverage (%), efficiency of collection of the generated wastewater (%) and coverage of municipal wastewater treatment (%).

To integrate all the information, the variables were normalized with respect to their minimum and maximum values ​​according to (Equation 1):

where:

Finally, the index of lack of access to water and sanitation services I_lawss is expressed by (Equation 2). The result is shown in Figure 1.

Figure 1 Lack of access to water and sanitation services index map by municipality I_lawss. 

Figure 1 identifies the municipalities with the greatest problems of supply and access to drinking water and sewerage services, both in urban and rural areas. The state of Oaxaca stands out, with deficiencies in 83% of its municipalities, that is, 476 municipalities are in regular, deficient or very deficient conditions. Due to its extension, Chihuahua has the largest area in a very deficient zone of this indicator (1- GWSAI), followed by Oaxaca and Durango.

Table 1 shows the number of municipalities divided in ranges of the indicator I_lawss. In summary, it can be said that at national level 27% of the population has problems of access to drinking water services, with very deficient, deficient or regular drinking water and sewerage services, and that this population is located in 1485 municipalities.

Drought Index (I_d)

In Mexico, the National Meteorological Service (Servicio Meteorológico Nacional or SMN in Spanish) is the official government agency in charge of providing meteorological (weather) and climatological information. To do this, it uses observation networks, such as automatic meteorological stations, synoptic observatories, radars, sounding radio stations and satellite image receiving stations. As part of its monitoring activities, SMN is in charge of detecting the current state and evolution of the drought (^{Conagua-SMN, 2018}) through the so called Drought Monitor in Mexico (MSM, from Monitor de Sequía in México in Spanish), which in turn is part of the North American Drought Monitor (NADM). MSM started to operate in 2002, within the activities of the NADM, but it was until 2014 that it acquired its national character, which issued drought maps on time scale different from the monthly, always based on the methodology by the United States Drought Monitor (USDM) and the NADM. As of February 2014, the issuance of the MSM is biweekly.

These indices are displayed in layers through a Geographic Information System (GIS), and the regions affected by drought are determined by consensus, according to the intensity scale that is common in the three countries (United States, Mexico and Canada) ranging from abnormally dry (D0), moderate drought (D1), severe drought (D2), extreme drought (D3) to exceptional drought (D4).

In this context, based on the MSM historical records of for each drought category, and on municipal data, a methodology was developed to prepare a municipal level national drought hazard map, which is described below in summary form. It is described extensively in ^{IMTA (2017)}, and ^{Ortega-Gaucin, De la Cruz, and Castellano (2018)}.

Municipal scale drought hazard was determined by calculating the probability of drought occurrence (relative frequency) using MSM records, which involved carrying out the following activities:

Equation (3) was used to calculate the relative frequency:

where:

The probability values oscillate in a closed range of [0, 1]. In order to classify these probabilities in percentiles with categories of equal size, the values were adjusted to a normal distribution and the result was multiplied by 100. Given that there are five categories of drought probability classification, each percentile range obeys 20% (Figure 2).

Figure 2 Drought index map by municipality I_d in the Mexican Republic. 

The state with the highest number of affected municipalities is Michoacán, where 70% of them have a very high drought, in addition to the northern states such as Baja California, Sonora, Coahuila and the southeastern states of Campeche, Quintana Roo and Yucatan. The state with the highest number of municipalities with high drought is Veracruz with 71 municipalities.

The state with the highest number of municipalities in the middle, low and very low range is Oaxaca, with 116 municipalities in the middle level, 108 municipalities in the low level and 323 municipalities in the very low level, that is to say 96% of the total number of municipalities in that state present some degree of drought.

Table 2 presents a breakdown of drought ranges, with number of municipalities, population and area in each range, highlighting that 17.9% of the municipalities (where 22.3% of the population live) is located in very high probability of drought areas.

Groundwater depletion index (I_gwd)

Groundwater is still considered by many as an invisible and inexhaustible resource. Actually, most aquifer systems have been subject to an intensive use never seen before, causing local problems (land subsidence) and regional problems (interception of regional flows, which affect interconnected aquifer units, and reduction in their natural discharge outlets), with irreversible damage (^{Custodio, 2000}).

In Mexico, it is necessary to understand the dynamics of groundwater, which is much slower than that of surface water. The phreatic and piezometric level aquifer monitoring is limited in space (systematic variation in the measurement points, absence of a spatial arrangement, of the measurement networks) and in time (variation at different times of the year, lack of long-term continuity). Since the 1970s, efforts have been made to establish monitoring networks in free, semi-confined and confined aquifers in the country, but data gaps in monitoring persist, which is why remote sensing is a technology that little by little has been integrated into the hydro-sciences, offering the advantage of monitoring also difficult access areas, with a wealth of systematic and continuous information.

The “Gravity Recovery and Climate Experiment” (GRACE) project allows groundwater volumes to be estimated from temporal variations in the gravity field. GRACE is an American-German project, carried out by the National Aeronautics and Space Administration (NASA) and the German Aerospace Center (DLR) by Deutsches Zentrum für Luft - und Raumfahrt), consisting of two twin satellites that orbit 450 km above sea level, 220 km apart. The spatial resolution of the information is approximately 200 000 km², applicable to hydrological basins (^{Wahr, Zlotnicki, & Velicogna, 2004}; ^{Scalon, Longuevergne, & Long, 2012}). GRACE generates temporal information of the gravitational field with a periodicity of 10 days to a month. GRACE L2 series information is available at: http://www2.csr.utexas.edu/grace/RL06.html.

The geoid or land mass density (^{Wahr, Zlotnick, & Velicogna, 1998}) is represented in the form of finite series of spherical harmonics, with which the total variation of water storage (VTAA) is calculated, expressed in units of water equivalent. The VTAA integrates changes in the storage of: ice-snow (CAHN), surface water (CAAS), soil water (CAS) and subsoil (CASUB), the latter being estimated with Equation (4):

The novelty of GRACE is the recording of the temporal variation of the gravitational field, translated in VTAA globally.

The objective of using the GRACE data and reducing them to obtain CASUB lies in obtaining the average depletion-recovery in the Mexican Republic, information that is used as groundwater index in the calculation of municipal water security indices.

Applying the standard methodology of the GRACE data, from the L2 series, monthly VTAA records were obtained for the period of time 2002 to 2015 and the CASUB value was obtained by applying Equation 4. The information from the CAHN, CAAS and CAS components was processed from 1° terrestrial resolution databases, on a monthly period, from 2002 to 2015.

It is noteworthy that the mass changes where the greatest variation is observed correspond to the seasons winter-summer and spring-autumn cycles, that is, there is a seasonal transition phenomenon. In the winter season are the greatest depletions and in summer the recoveries, which are associated with the effect of precipitation. This phenomenon is typical in arid and semi-arid areas. In the spring-autumn season there is an intermediate process where the total value of water storage is recovered or reduced.

In general, it is identified that in the north of the country a total value of water storage between stable and depletion is maintained. In the central region, represented mostly by the Lerma-Santiago Pacifico, Golfo Norte, Aguas del Valle de México, Balsas, Pacifico Sur, Golfo Centro and Frontera Sur basin councils, the greatest water mass variations occur. In the Yucatan Peninsula it tends to be stable, although the latter must be taken with reserve since the local processes are the dominant ones, and are hidden when analyzed with the spherical harmonics of minor orders and degrees.

The average CASUB value for July from the years 2002 to 2015 shows that infiltration and recharge extend to almost the entire coastal area of ​​the Gulf of Mexico, including the Yucatan Peninsula and continue towards the Valley of Mexico basin and the head of the Panuco River basin. Recharge and infiltration are observed on the coasts of the Pacific Ocean during the month of July, mainly for the middle zone of the Baja California Peninsula, towards the Vizcaino desert area.

In August, the conditions existing in July increase, but they extend to the areas of the Balsas River basin, the North Pacific coastal basin between the cities of Hermosillo and Mazatlán, the head of the Lerma River basin and the eastern portion of the Central Basins.

October is similar to September, in the distribution, however, it is more intense towards the central portion of Mexico, mainly towards the headwaters of Valle de México, Rio Lerma, Río Panuco and Balsas basins; coasts of Veracruz (Xalapa) as well as in the portion of the Yucatan Peninsula and the Oaxaca coasts.

By the end of the rainy season, in November, the groundwater storage effects are manifested in the same way, with a decrease or progressive attenuation of said areas for December.

Groundwater storage was calculated month by month, for the time period 2002-2015, and the monthly and annual average values were subsequently estimated. The result is the average groundwater storage, over a period of 13 years, where groundwater storage is considered as recovery if positive, and depletion if negative. To obtain the storage values ​​by municipality, an interpolation was performed with the results of Figure 3 and the municipal division. Normalization of the resulting values was applied after that.

Figure 3 Average value of CASUB (the change of water storage in the subsoil) in the Mexican Republic, for time period 2002-2015. 

The normalization method applied to the depletion/recovery values, on 0 to 1 scale, was Min-Max (^{OECD, 2008}). The classification of data to determine the best categorization of the values, according to the breaks or gaps that naturally exist between them, was carried out by maximizing the variance between classes and minimizing the variance within classes (^{Jenks, 1967}), using the QGIS software, see graphic results on the depletion map in Figure 4.

Figure 4 Groundwater depletion hazard index map by municipality I_gwd in the Mexican Republic. 

The State of Jalisco has the largest number of municipalities with very high groundwater depletion (85), representing 68% of the total of its municipalities. In a high and medium range of depletion is the state of Oaxaca with 115 municipalities in each case, although it also has 200 municipalities in acceptable conditions. The state of Veracruz has 102 of municipalities with no depletion (48%).

At national level, there are 519 municipalities with very high depletion (21%), with a population of 22 520 521 inhabitants and an area of ​​369 377.33 km². 431 municipalities, where 26 504 321 people live in an area of ​​356 627.33 km² have high groundwater depletion. The municipalities in medium range of depletion are 397. Finally, 1 109 municipalities (45.15% of the total of municipalities), where 39 300 516 people live, have very low depletion (Table 3).

Flood and extreme meteorological event index (I_feme)

The National Water Information System (SINA), belonging to CONAGUA, publishes information, at municipal level, on declarations of emergency due to hydro-meteorological phenomena, issued from 2000 to 2016. In Mexico, there are procedures for issuing declarations for these phenomena, in which the National Center for Disaster Prevention (CENAPRED) participates. The extracted information contains data on cyclones, rain, floods, low temperatures, snow and hail, and tornadoes (^{Conagua-Sina, 2016}).

The total number of emergency declarations by municipality was taken, from the information for that period, which includes cyclones, heavy rains, floods, low temperatures, strong winds, snow-frost and tornadoes. This number was divided among the days of a year to obtain an annual value of declarations by municipality. This information was normalized by using a normal distribution function (Equation 5), with mean μ and variance σ² obtained from the annual declaratory value series per municipality.

The classification of data to determine the best categorization of the values, according to the breaks or gaps that naturally exist between the data, was achieved by maximizing the variance between classes and minimizing the variance within the classes (^{Jenks, 1967}), using the QGIS software.

Figure 5 shows the resulting hydro-meteorological phenomena hazard map. The state with the highest recurrence (very high level) of extreme meteorological phenomena is Veracruz, with 70% of its municipalities (149) in that range. With a high, medium, low and very low level is the state of Oaxaca, with 96% of its municipalities ravaged by some extreme weather.

Figure 5 Flood and extreme meteorological event index map (I_feme). 

The affected population in the very high range is 22 676 963.00 people, located in an area of 529 637.77 km², of 330 affected municipalities. In the high range are 359 municipalities, with 14 637 033 inhabitants, located in 429 908.29 km². The low and very low ranges, with a minimum frequency of these events, includes 1 426 municipalities with 64 642 429 inhabitants (Table 4).

Hazard index (I_h)

In order to integrate these indices into one index, they were normalized, on 0 to 1 scale, and the hazard index (I_h) was calculated by Equation 6. The hazard index map resulting from the integration of these four indices is presented in Figure 6.

Figure 6 Municipal hazard index I_h.