• Alhamed, A., , S. Lakshmivarahan, , and D. J. Stensrud, 2002: Cluster analysis of multimodel ensemble data from SAMEX. Mon. Wea. Rev., 130, 226256, doi:10.1175/1520-0493(2002)130<0226:CAOMED>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Allen, R. G., , M. Smith, , L. S. Pereira, , and A. Perrier, 1994: An update for the calculation of reference evapotranspiration. Int. Comm. Irrig. Drain. Bull., 43, 3592.

    • Search Google Scholar
    • Export Citation
  • Allen, R. G., , L. S. Pereira, , D. Raes, , and M. Smith, 1998: Crop evapotranspiration: Guidelines for computing crop water requirements. FAO Irrigation and Drainage Paper 56, 300 pp. [Available online at http://www.fao.org/docrep/x0490e/x0490e00.htm.]

  • Brito-Castillo, L., 2012: Regional patterns of trends in long-term precipitation and stream flow observations: Singularities in a changing climate in Mexico. Greenhouse Gases: Emission, Measurement, and Management, G. Liu, Ed., In Tech, 387–412, doi:10.5772/32804.

  • Brito-Castillo, L., , E. R. Vivoni, , D. J. Gochis, , A. Filonov, , I. Tereshchenko, , and C. Monzon, 2010: An anomaly in the occurrence of the month of maximum precipitation distribution in northwest Mexico. J. Arid Environ., 74, 531539, doi:10.1016/j.jaridenv.2009.10.014.

    • Search Google Scholar
    • Export Citation
  • Cavazos, T., 1999: Large-scale circulation anomalies conducive to extreme precipitation events and derivation of daily rainfall in northeastern Mexico and southeastern Texas. J. Climate, 12, 15061523, doi:10.1175/1520-0442(1999)012<1506:LSCACT>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Comrie, A. C., , and E. C. Glenn, 1998: Principal components-based regionalization of precipitation regimes across the southwest United States and northern Mexico, with an application to monsoon precipitation variability. Climate Res., 10, 201215, doi:10.3354/cr010201.

    • Search Google Scholar
    • Export Citation
  • Cook, B. I., , R. Seager, , and R. Miller, 2011: Atmospheric circulation anomalies during two persistent North American droughts: 1932–1939 and 1948–1957. Climate Dyn., 36, 23392355, doi:10.1007/s00382-010-0807-1.

    • Search Google Scholar
    • Export Citation
  • Dai, A., , K. E. Trenberth, , and T. Qian, 2004: A global data set of Palmer drought severity index for 1870–2002: Relationship with soil moisture and effects of surface warming. J. Hydrometeor., 5, 11171130, doi:10.1175/JHM-386.1.

    • Search Google Scholar
    • Export Citation
  • Douglas, M. W., , R. A. Maddox, , K. Howard, , and S. Reyes, 1993: The Mexican monsoon. J. Climate, 6, 16651677, doi:10.1175/1520-0442(1993)006<1665:TMM>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Ekstrom, M., , P. D. Jones, , H. Fowler, , G. Lenderink, , T. A. Buishand, , and D. Conway, 2007: Regional climate model data used within the SWURVE project—1: projected changes in seasonal patterns and estimation of PET. Hydrol. Earth Syst. Sci., 11, 10691083, doi:10.5194/hess-11-1069-2007.

    • Search Google Scholar
    • Export Citation
  • Garcia, E., , R. Vidal, , and M. E. Hernandez, 1990: Las regiones climáticas de México (The climatic regions of Mexico). Atlas Nacional de México, Vol. 2, A. Garcia de Fuentes, Ed., Universidad Nacional Autónoma de México, Instituto de Geografia, 61 pp.

  • Giddings, L. S. M., , B. M. Rutherford, , and A. Maarouf, 2005: Standardized precipitation index zones for Mexico. Atmósfera, 18, 3356.

  • Giovannettone, J. P., , and A. P. Barros, 2008: A remote sensing survey of the role of the landform on the organization of orographic precipitation in central and southern Mexico. J. Hydrometeor., 9, 12671283, doi:10.1175/2008JHM947.1.

    • Search Google Scholar
    • Export Citation
  • Hare, F. K., 1993: Climate variations, drought and desertification. World Meteorological Organization 653, 45 pp.

  • Harris, I., , P. D. Jones, , T. J. Osborn, , and D. H. Lister, 2013: Updated high-resolution grids of monthly climatic observations: The CRU TS3.10 dataset. Int. J. Climatol., 34, 623–642, doi:10.1002/joc.3711.

  • Hulme, M., , R. March, , and P. D. Jones, 1992: Global changes in a humidity index between 1931–60 and 1961–90. Climate Res., 2, 122, doi:10.3354/cr002001.

    • Search Google Scholar
    • Export Citation
  • IPCC, 2007: Climate Change 2007: The Physical Science Basis. Cambridge University Press, 996 pp.

  • Jauregui, E., 2003: Climatology of land falling hurricanes and tropical storms in Mexico. Atmósfera, 16, 193204.

  • Kushnir, Y., , R. Seager, , M. Ting, , N. Naik, , and J. Nakamura, 2010: Mechanisms of tropical Atlantic SST influence on North American precipitation variability. J. Climate, 23, 56105628, doi:10.1175/2010JCLI3172.1.

    • Search Google Scholar
    • Export Citation
  • Lizárraga-Celaya, C., , C. J. Watts, , J. C. Rodriguez, , J. Garatuza-Payan, , R. L. Scott, , and J. Sáiz-Hernández, 2010: Spatio-temporal variation in surface characteristics over the North American monsoon region. J. Arid Environ., 74, 540548, doi:10.1016/j.jaridenv.2009.09.027.

    • Search Google Scholar
    • Export Citation
  • Lu, J., , G. A. Vecchi, , and T. Reichler, 2007: Expansion of the Hadley cell under global warming. Geophys. Res. Lett., 34, L06805, doi:10.1029/2006GL028443.

    • Search Google Scholar
    • Export Citation
  • Magaña, V., , J. Amador, , and S. Medina, 1999: The midsummer drought over Mexico and Central America. J. Climate, 12, 15771588, doi:10.1175/1520-0442(1999)012<1577:TMDOMA>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Méndez, M., , and V. Magaña, 2010: Regional aspects of prolonged meteorological droughts over Mexico and Central America. J. Climate, 23, 11751188, doi:10.1175/2009JCLI3080.1.

    • Search Google Scholar
    • Export Citation
  • Mendez-Barroso, L. A., , E. R. Vivoni, , Ch. J. Watts, , and J. C. Rodriguez, 2009: Seasonal and interannual relations between precipitation, surface soil moisture and vegetation dynamics in the North American monsoon region. J. Hydrol., 377, 5970, doi:10.1016/j.jhydrol.2009.08.009.

    • Search Google Scholar
    • Export Citation
  • Pavia, E. G., , and F. Graef, 2002: The recent rainfall climatology of the Mediterranean Californias. J. Climate, 15, 26972701, doi:10.1175/1520-0442(2002)015<2697:TRRCOT>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Penman, H. L., 1948: Natural evaporation from open water, bare soil and grass. Proc. Roy. Soc., 193A, 120145, doi:10.1098/rspa.1948.0037.

    • Search Google Scholar
    • Export Citation
  • Reich, R. M., , B. C. Aguirre-Bravo, , and V. A. Bravo, 2008: New approach for modeling climatic data with applications in modeling tree species distributions in the states of Jalisco and Colima, Mexico. J. Arid Environ., 72, 13431357, doi:10.1016/j.jaridenv.2008.02.004.

    • Search Google Scholar
    • Export Citation
  • Romesburg, C. H., 1984: Cluster Analysis for Researchers. Life Time Learning, 334 pp.

  • Scheff, J., , and D. M. W. Frierson, 2014: Scaling potential evapotranspiration with greenhouse warming. J. Climate, 27, 15391558, doi:10.1175/JCLI-D-13-00233.1.

    • Search Google Scholar
    • Export Citation
  • Seager, R., 2007: The turn of the century North American drought: Global context, dynamics, and past analogs. J. Climate, 20, 55275552, doi:10.1175/2007JCLI1529.1.

    • Search Google Scholar
    • Export Citation
  • Seager, R., , N. Naik, , M. Ting, , M. A. Cane, , N. Harnik, , and Y. Kushnir, 2010: Adjustment of the atmospheric circulation to tropical Pacific SST anomalies: Variability of transient eddy propagation in the Pacific–North America sector. Quart. J. Roy. Meteor. Soc., 136, 277296, doi:10.1002/qj.588.

    • Search Google Scholar
    • Export Citation
  • Shein, K. A., 2006: State of the climate in 2005. Bull. Amer. Meteor. Soc., 87 (6), S1–S102, doi:10.1175/BAMS-87-6-shein.

  • Stahle, D. W., and Coauthors, 2009: Early 21st-century drought in Mexico. Eos, Trans. Amer. Geophys. Union, 90, 8990, doi:10.1029/2009EO110001.

    • Search Google Scholar
    • Export Citation
  • Szekely, G. J., , and M. L. Rizzo, 2005: Hierarchical clustering via joint between-within distances: Extending Ward’s minimum variance method. J. Classif., 22, 151183, doi:10.1007/s00357-005-0012-9.

    • Search Google Scholar
    • Export Citation
  • Tereshchenko, I., , A. N. Zolotkrylin, , T. B. Titkova, , L. Brito-Castillo, , and C. O. Monzón, 2012: Seasonal variation of surface temperature-modulating factors in the Sonoran Desert in northwestern Mexico. J. Appl. Meteor. Climatol., 51, 15191530, doi:10.1175/JAMC-D-11-0160.1.

    • Search Google Scholar
    • Export Citation
  • Thornthwaite, C. W., 1948: An approach towards a rational classification of climate. Geogr. Rev., 38, 5594, doi:10.2307/210739.

  • Tian, D., , and C. Martinez, 2014: The GEFS-based daily reference evapotranspiration (ETo) forecast and its implication for water management in the southeastern United States. J. Hydrometeor., 15, 11521165, doi:10.1175/JHM-D-13-0119.1.

    • Search Google Scholar
    • Export Citation
  • Tucker, C. J., , J. E. Pinzon, , M. E. Brown, , D. Slayback, , E. W. Pak, , R. Mahoney, , E. Vermote, , and N. El Saleous, 2005: An extended AVHRR 8-km NDVI dataset compatible with MODIS and SPOT vegetation NDVI data. Int. J. Remote Sens., 26, 44854498, doi:10.1080/01431160500168686.

    • Search Google Scholar
    • Export Citation
  • UNCCD, 1994: United Nations convention to combat desertification in those countries experiencing serious drought and desertification, particularly in Africa. United Nations Treaty Series, Vol. 1954, 3. [Available online at https://treaties.un.org/pages/ViewDetails.aspx?src=TREATY&mtdsg_no=XXVII-10&chapter=27&lang=en.]

  • UNEP, 1992: World Atlas of Desertification. Edward Arnold, 69 pp.

  • UNESCO, 1979: Map of the world distribution of arid regions. UNESCO Man and the Biosphere (MAB) Tech. Note 7, 54 pp.

  • University of East Anglia Climatic Research Unit, 2013: CRU TS3.21: Climatic Research Unit (CRU) time series (TS) version 3.21 of high-resolution gridded data of month-by-month variation in climate (January 1901–December 2012). NCAS British Atmospheric Data Centre, accessed 2013–July 2014, doi:10.5285/D0E1585D-3417-485F-87AE-4FCECF10A992.

  • Vivoni, E. R., , H. A. Moreno, , G. Mascaro, , J. C. Rodriguez, , C. J. Watts, , J. Garatuza-Payan, , and R. L. Scott, 2008: Observed relation between evapotranspiration and soil moisture in the North American monsoon region. Geophys. Res. Lett., 35, L22403, doi:10.1029/2008GL036001.

    • Search Google Scholar
    • Export Citation
  • Watts, C. J., , R. L. Scott, , J. Garatuza-Payan, , J. C. Rodriguez, , J. H. Pruegger, , W. P. Kustas, , and M. Douglas, 2007: Changes in vegetation condition and surface fluxes during NAME 2004. J. Climate, 20, 18101820, doi:10.1175/JCLI4088.1.

    • Search Google Scholar
    • Export Citation
  • Wilks, D. S., 1995: Statistical Methods in the Atmospheric Sciences: An Introduction. Academic Press, 467 pp.

  • Zolotokrylin, A. N., 2003: Climatic Desertification (in Russian). Nauka, 246 pp.

  • View in gallery

    Location of meteorological stations in Mexico spanning more than 50 years of data (1951–2001).

  • View in gallery

    Long-term mean annual AI in Mexico during 1951–2001.

  • View in gallery

    Changes of the PDSI in Mexico in the period 1951–2001.

  • View in gallery

    Composite difference map of AI in (a) 1951–57, (b) 1960–65, and (c) 1994–2001 relative to 1961–90. Areas with statistically significant differences are hatched.

  • View in gallery

    (a) Clustering results, (b) regionalization (clusters) of the AI during the decade 1951–2001, and (c) interannual changes in the AI with 11-yr moving-average smoothing curves. The colors and numbers in (b) and (a) correspond to the color lines and numbers in (c). Inset in (a) displays the link between the number of clusters and average within-cluster correlation. Inset in (b) shows the correlation coefficients between the center of each cluster and its members.

  • View in gallery

    (a) Location of weather stations with more than 40 yr of observations that display the month of the main peak summer rainfall (between May and October) in Mexico. (b) As in (a), but the symbols display the month when a second peak occurs.

  • View in gallery

    The AI (contours) and AVHRR NDVI values (colors) in 1982–2001: (a) annual mean; (b) May–September mean.

  • View in gallery

    Linear regression between AVHRR NDVI annual time series and average values of AI from 1982 to 2001. Points associated with low vegetation on dry lands are bordered by the red rectangle.

  • View in gallery

    Changes in the AI and the AVHRR NDVI anomalies in 1992–2001 relative to 1982–91: (a) AI, (b) annual cycle of NDVI, and (c) NDVI (May–September). Areas with statistically significant differences are hatched.

  • View in gallery

    The 0.75 contour line of the AI during (a) 1982–91 and (b) 1992–2001. Areas with AI values lower than 1 std dev (based on 1951–80 data) areas are in green.

  • View in gallery

    For (a) EOF 1 and (b) EOF 2, (left) the leading EOF modes of the annual AI in the period 1951–2001 and (right) their temporal variability (blue line) and 11-yr moving-average smoothing curves (red line). The percentages of explained variance for each mode are shown at the top of the right panels.

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 31 31 13
PDF Downloads 12 12 6

Changes in Aridity across Mexico in the Second Half of the Twentieth Century

View More View Less
  • 1 Departamento de Física, Centro Universitario de Ciencias Exactas e Ingeniería, Universidad de Guadalajara, Guadalajara, Jalisco, Mexico
  • 2 Climatology Laboratory, Institute of Geography, Russian Academy of Sciences, Moscow, Russia
  • 3 Departamento de Física, Centro Universitario de Ciencias Exactas e Ingeniería, Universidad de Guadalajara, Guadalajara, Jalisco, Mexico
  • 4 Centro de Investigaciones Biologicas del Noroeste, Guaymas, Sonora, Mexico
  • 5 Climatology Laboratory, Institute of Geography, Russian Academy of Sciences, Moscow, Russia
© Get Permissions
Full access

Abstract

Six regions in Mexico, with typical interannual changes in the aridity index, have been defined by the 1951–2001 meteorological dataset. Peak months of rainfall differ within the regions. Most of the land in the Mexican terrain has had a slow aridization since the early 1980s. The decline in the aridity index in the early 1950s and late 1990s was caused by droughts in the area. The distinctive features of the aridization of Mexican dry lands are characterized by steady and extensive droughts during 1948–57, 1960–65, and 1994–2003 in the second half of the twentieth century. During the drought of 1951–57 substantial aridization in most of the dry lands was observed, including the Sierra Madre Occidental, Sierra Madre Oriental, and Mexican Altiplano. Aridization of dry lands during the drought in 1960–65 affected mostly the southern part of the Mexican Altiplano, the Sierra Madre del Sur, and the Yucatán Peninsula. For the drought in the 1990s, one special feature of the aridization was its propagation primarily beyond the Mexican Altiplano. Increased aridization of dry lands caused by long-term droughts during the last decade of the twentieth century did not result in a sizeable shift of the southern boundary of the dry lands. The only exception is the southern boundary (aridity index = 0.75) in the state of Sinaloa. In this area, the boundary moved southward and aridization intensified. The results obtained here can be used in studies of possible anthropogenic impact on the drought of the twentieth century’s last decade in Mexico, which includes changes in land use.

Corresponding author address: Luis Brito-Castillo, Centro de Investigaciones Biológicas del Noroeste (CIBNOR), Unidad Sonora, Campus Guaymas, Km. 2.35 camino al tular, estero de Bacochibampo, Guaymas, Sonora 85454, Mexico. E-mail: lbrito04@cibnor.mx

Abstract

Six regions in Mexico, with typical interannual changes in the aridity index, have been defined by the 1951–2001 meteorological dataset. Peak months of rainfall differ within the regions. Most of the land in the Mexican terrain has had a slow aridization since the early 1980s. The decline in the aridity index in the early 1950s and late 1990s was caused by droughts in the area. The distinctive features of the aridization of Mexican dry lands are characterized by steady and extensive droughts during 1948–57, 1960–65, and 1994–2003 in the second half of the twentieth century. During the drought of 1951–57 substantial aridization in most of the dry lands was observed, including the Sierra Madre Occidental, Sierra Madre Oriental, and Mexican Altiplano. Aridization of dry lands during the drought in 1960–65 affected mostly the southern part of the Mexican Altiplano, the Sierra Madre del Sur, and the Yucatán Peninsula. For the drought in the 1990s, one special feature of the aridization was its propagation primarily beyond the Mexican Altiplano. Increased aridization of dry lands caused by long-term droughts during the last decade of the twentieth century did not result in a sizeable shift of the southern boundary of the dry lands. The only exception is the southern boundary (aridity index = 0.75) in the state of Sinaloa. In this area, the boundary moved southward and aridization intensified. The results obtained here can be used in studies of possible anthropogenic impact on the drought of the twentieth century’s last decade in Mexico, which includes changes in land use.

Corresponding author address: Luis Brito-Castillo, Centro de Investigaciones Biológicas del Noroeste (CIBNOR), Unidad Sonora, Campus Guaymas, Km. 2.35 camino al tular, estero de Bacochibampo, Guaymas, Sonora 85454, Mexico. E-mail: lbrito04@cibnor.mx

1. Introduction

The ratio of average annual precipitation to annual potential evapotranspiration (P/PET) is useful for characterizing regionalization of subcontinental-scale areas, such as Mexico, based on aridity. The P/PET ratio rises with increasing humidification of the area. Therefore, the P/PET index is a measure of moisture surplus or deficit (Hulme et al. 1992). The P/PET ratio is also used for classifying dry lands and is referred to as the aridity index (Hare 1993). The most recent classification of arid zones was published by the United Nations Environment Programme (UNEP 1992) as an index of desertification for the World Atlas of Desertification. Hereinafter we refer to the P/PET ratio as the aridity index (AI). When the index decreases, the dryness of the area is more intense.

Of all the approaches for estimating potential evapotranspiration (Thornthwaite 1948; Penman 1948), UNEP (1992) recommends the most straightforward method, in which potential evapotranspiration is a function of temperature and latitude, which was first defined by Thornthwaite (1948). The UNEP approach was endorsed by the United Nations Convention to Combat Desertification (UNCCD 1994). According to the UNCCD, the dry lands are areas where the aridity index ranges from 0.05 to 0.65 (UNCCD 1994).

However, there are studies claiming that the Penman method is more appropriate for the assessment of potential evapotranspiration (Tian and Martinez 2014; Scheff and Frierson 2014). Therefore in this study the Penman equation has been used for estimating PET. Four main aridity zones were delimited by the United Nations Educational, Scientific and Cultural Organization (UNESCO 1979): hyperarid (<0.03), arid lands (0.03–0.2), semiarid lands (0.2–0.5), and dry subhumid lands (0.5–0.75). Most Mexican dry lands are either impacted by desertification or under threat of desertification. In recent decades, anthropogenic-based desertification has become more important.

Global climate forecasts predict aridization of subtropical regions and, as a consequence, the probable intensification of the climate component of desertification (IPCC 2007). The most sensitive regions are arid lands of Mexico including the vast Sonoran and Chihuahuan Deserts. The purpose of this study is to determine the areas of interannual changes in aridity index patterns within Mexico in the second half of the twentieth century and to describe the features of these patterns within the defined zones. Regionalization of aridity index is necessary to answer important questions: what are the aridity index trends across Mexico, and does the aridity of dry lands increase? In addition, the variability of dry lands whose border is defined as a contour line with AI = 0.75 during the last decade of the twentieth century (which is a period of intensive global warming; IPCC 2007) relative to the preceding decade (1982–91) was investigated. The results obtained are substantiated by comparisons of the AI trends during the last two decades of the twentieth century with interdecadal variability of satellite-sensed vegetation index data, using the normalized difference vegetation index (NDVI). Threshold values of the NDVI may be used as an indicator of switches in surface temperature-modulating factors within a seasonal cycle (Tereshchenko et al. 2012; Zolotokrylin 2003).

The NDVI is a remotely sensed satellite-based index that measures the “greenness” of the vegetation cover. Its use and application simplify the monitoring of desertification (Watts et al. 2007; Vivoni et al. 2008; Mendez-Barroso et al. 2009; Lizárraga-Celaya et al. 2010).

2. Study area, data, and methods

Dry lands cover a large part of Mexico. They extend from the north of the U.S.–Mexico border and gradually narrow in the central part of Mexico near 22°N. More than 350 meteorological stations covering 51 yr (1951–2001) were selected, but we used only 237 station data with less than 25% missing values (Fig. 1). Daily precipitation and temperature records were provided by the National Meteorological Service (NMS) of Mexico. Data quality was verified before calculations through direct inspection of each series. When an individual record was found to be extremely high or extremely low without replication in surrounding stations, the record was considered a gap. Negative values were removed from precipitation records and replaced by gaps. Such inspection is routinely done in NMS to prevent possible errors in calculations. The monthly mean temperatures and monthly precipitation sums for each station were computed from daily temperature means and daily precipitation sums.

Fig. 1.
Fig. 1.

Location of meteorological stations in Mexico spanning more than 50 years of data (1951–2001).

Citation: Journal of Applied Meteorology and Climatology 54, 10; 10.1175/JAMC-D-14-0207.1

The AI was calculated as AI = P/PET, where P is the average annual precipitation (computed from monthly precipitation sums) in millimeters per year and PET is the annual potential evapotranspiration in millimeters per year.

The monthly PET and total precipitation data of gridded resolution 0.5° × 0.5° were obtained from the Climatic Research Unit (CRU) of the University of East Anglia global monthly dataset CRU TS 3.21 (University of East Anglia Climatic Research Unit 2013) and were used to analyze the aridity index variations. The potential evapotranspiration on the continental and subcontinental scale is obtained from the Food and Agriculture Organization (FAO) version (Allen et al. 1998), which is one of the most widely known and available methods. It is calculated based on the FAO grass reference evapotranspiration equation [Ekstrom et al. (2007), which is based on Allen et al. (1994)]:
e1
where ET0 is grass reference evapotranspiration (mm day−1), is net radiation at crop surface (MJ m−2 day−1), G is soil heat flux (MJ m−2 day−1), T is mean daily temperature at 2-m height (°C), is wind speed measured at 2-m height (m s−1), is a vapor pressure deficit for measurement at 2-m height (kPa), Δ is the slope of the vapor pressure curve (kPa °C−1), and γ is the psychrometric constant (kPa °C−1).

Equation (1) is a variation of the Penman–Monteith method, which uses the gridded daily mean temperature, monthly average daily minimum temperature, monthly average daily maximum temperature, vapor pressure, and cloud cover data (Harris et al. 2013). The average monthly values of the wind speed for the period of 1961–90 were used as a source data. Long-term means of annual values of AI, P, and PET were calculated for 1951–2001, 1951–80, 1961–90, 1982–91, 1982–2001, and 1992–2001. The t test for independent samples by groups was applied for determination of statistical significance of differences.

In general, any trend in climate variables identified for the country as a whole can obscure regional trends. In this case a regionalization reveals dissimilarities clarifying the regional behavior. The spatial distribution of the areas generated from the hierarchical cluster analysis (HCA) algorithm (Romesburg 1984; Szekely and Rizzo 2005) based on the 1951–2001 time series of AI is in good agreement with the aridity zones defined according to the Penman (1948) aridity index (UNESCO 1979). For the regionalization of Mexico, based on quasi-homogeneous changes in humidification patterns, we clustered the interannual changes in AI during the 1951–2001 half century. Regionalization of AI was done with the SPSS statistical-analysis software package.

HCA is a statistical technique for grouping objects together using a selected similarity measure. HCA algorithm has been described in detail (e.g., Alhamed et al. 2002). In this study the Pearson correlation was chosen as a measure of similarity among the AI time series.

The high quantity of clusters makes the interpretation of results too complex, therefore our goal is to minimize cluster count, keeping an acceptable level of coherence. To identify acceptable cluster count, we used the weighted average (Waverage) method (average linkage within groups) and estimated Pearson correlation (as a metric) between the center of the cluster and all its members, starting from the case when all nodes fall into one single cluster. The center of the cluster is calculated as the mean time series derived from all AI time series/members of the cluster. We took into consideration that the correlations between each cluster center and its members were significant. It is also important that the spatial distribution of clusters would display the features of the territory of Mexico (especially orographic). This criterion was accepted as a within-group linkage of desired level. The application of the kriging interpolation algorithm to the clustering results (i.e., the georeferenced points of each cluster) resulted in the smooth boundaries of regions. In addition, the interannual changes of the AI and 11-yr moving-average smoothing curves were generated.

The regionalization process of Mexico was carried out by other authors as well. Garcia et al. (1990) compiled charts of zones with different precipitation regimes and temperature patterns. Comrie and Glenn (1998) made a regionalization of precipitation regimes across northern Mexico, based on principal components analysis, which allows them to study precipitation patterns, in particular, monsoon subregional precipitation variability. Giddings et al. (2005) produced zoning based on homogeneous precipitation patterns, using the standardized precipitation index. Reich et al. (2008) applied cluster analysis to temperature, precipitation and evaporation data to delineate climatic humidification zones for the tropical, temperate, and semiarid forests in the states of Jalisco and Colima.

The method of empirical orthogonal function (EOF) analysis was used to study the spatial–temporal structure of long-term variations in the aridity index. The rotated EOF modes were obtained by the varimax method. The objective of the EOF analysis is to explain the pattern of correlations within a set of observed variables (Wilks 1995).

The self-calibrated Palmer drought severity index (PDSI) data of gridded resolution 2.5° × 2.5° were obtained from the global monthly dataset of the National Center for Atmospheric Research (Dai et al. 2004) and were used to analyze the dry periods. The negative values indicate dry spells.

The circulation indices characterizing the El Niño intensity and the anomalies of the tropical Atlantic Ocean sea surface temperature were used to investigate the influence of the global atmospheric circulation on the interannual variability of the aridity index. The monthly data of the Niño SST indices and the monthly data of Atlantic tropical SST indices were obtained from the archive of the Climate Prediction Center (http://www.cpc.ncep.noaa.gov/). For indices of tropical Pacific and tropical Atlantic SST variability, we used the Niño indices (Niño-1+2, Niño-3, Niño-4), defined as the normalized SST anomalies in the region lying between 160°E and 80°W and between 10°S and 5°N and the normalized SST anomalies over the tropical Atlantic [North Atlantic tropical (NATL) SST index and South Atlantic tropical (SATL) SST index] between 60°W and 10°E and between 20°S and 20°N.

The NDVI data are frequently used in analyzing changes in greenness or land surface phenology. We used the 8-km, 10-day (decade) composites of the Pathfinder AVHRR Land (PAL) NDVI data from the period 1982–2001 (Tucker et al. 2005). PAL data are a widely used standard dataset. Data from four NOAA AVHRR sensors (NOAA-7, NOAA-9, NOAA-11, and NOAA-14) constitute the PAL NDVI dataset from the period 1981–2001. The mean NDVI was computed for the whole year and the monsoon period (May–September) for 1982–2001, 1982–91, and 1992–2001.

3. Results

a. Long-term patterns of the aridity index across Mexico (1951–2001)

Dry lands constitute a continuous array in northern Mexico (Fig. 2). To the south of this area, separate arid regions are found mostly near the Pacific coast. Arid lands (0.0 AI < 0.2) occupy a significant part of the Baja California Peninsula and a small portion of Sonora; semiarid lands (0.2 AI < 0.5) on the Mexican Altiplano extend almost to 23°N. The irregular strip of dry subhumid lands (0.5 AI < 0.75) skirts the semiarid lands from the south, whereby the strip becomes noticeably wider in the western and eastern mountains of Sierra Madre Occidental and in coastal lowlands along the west coast. In the northern part of the Sierra Madre Oriental and northeast of it, where annual precipitation is higher, dry subhumid lands are surrounded by semiarid lands, while subhumid (0.75 AI < 0.80) and most other subhumid (0.8 AI < 1.0) lands are found in the highlands and along the coasts of Nayarit and Jalisco. Humid (1.0 AI < 1.2) and wet humid (≥1.2) lands are located mainly along the Gulf of Mexico coast and in southern Mexico (Fig. 2).

Fig. 2.
Fig. 2.

Long-term mean annual AI in Mexico during 1951–2001.

Citation: Journal of Applied Meteorology and Climatology 54, 10; 10.1175/JAMC-D-14-0207.1

Long-term droughts across Mexico are correlated with El Niño–Southern Oscillation (ENSO) and SST anomalies (Kushnir et al. 2010; Seager et al. 2010). Three droughts, 1948–57, 1960–65 and 1994–2003 (Stahle et al. 2009; Cook et al. 2011), are identified according to the PDSI data (Fig. 3 shows the period of 1951–2001 only). Negative AI anomalies are compared with the norm of 1961–90 occurrences not only within the dry lands boundaries but also outside these boundaries in the southern Mexico (Figs. 4a–c). There is a wet period during 1958–59 between the first two droughts (Fig. 3). The drought intensity in the period of 1960–65 is significantly weaker than the drought of the 1950s (excluding the southern half of Mexico; Figs. 3 and 4a,b). During the observation of the drought in 1951–57, substantial aridization in most of the dry lands is observed, including the Sierra Madre Occidental, Sierra Madre Oriental, and Mexican Altiplano (Fig. 4a). Aridization of dry lands during the drought in 1960–65 affects mostly the southern part of the Mexican Altiplano, the Sierra Madre del Sur, and the Yucatán Peninsula (Fig. 4b). During the drought in the 1990s, one special feature of the aridization is its propagation primarily beyond the Mexican Altiplano (Fig. 4c).

Fig. 3.
Fig. 3.

Changes of the PDSI in Mexico in the period 1951–2001.

Citation: Journal of Applied Meteorology and Climatology 54, 10; 10.1175/JAMC-D-14-0207.1

Fig. 4.
Fig. 4.

Composite difference map of AI in (a) 1951–57, (b) 1960–65, and (c) 1994–2001 relative to 1961–90. Areas with statistically significant differences are hatched.

Citation: Journal of Applied Meteorology and Climatology 54, 10; 10.1175/JAMC-D-14-0207.1

b. Regionalization of interannual changes in the aridity index during 1951–2001

Cluster analysis has been used since the 1950s and has been established as a reliable method of mathematical statistics, which is why we use this method in this work. Figure 5a displays the clustering results. The inset in Fig. 5a shows the increase of the average within-cluster correlation coefficient with the increase in the number of clusters. Note that correlation increases rapidly up to six clusters and then slows down when the number of clusters is larger than six. Therefore, we quite arbitrarily choose six clusters because the behavior of the correlation and the spatial connection, and also taking into account the climatic and orographic features of the territory. This way, modeling (HCA) of the spatial distribution of different types of quasi-homogeneous interannual changes in AI allowed us to characterize six basic cluster regions (Figs. 5a,b). Thus, nodes with the same interannual changes of AI were grouped together into single cluster (Fig. 5a). The result of regionalization was plotted on a map (Fig. 5b). The interannual changes of AI in these regions were fitted by 11-yr moving averages (Fig. 5c). The first region included the dry lands in northwestern Mexico and the northern part of the Mexican Altiplano. The region received rain from the monsoon pattern (Comrie and Glenn 1998; Douglas et al. 1993), except in the northwestern portion of the Baja California Peninsula, where most of the rainfall occurred in winter (Pavia and Graef 2002). In the former area, precipitation peaked in mid- and late summer and greatly decreased during the rest of the year (Fig. 6a). One distinctive feature of this region was a small rise in the AI during 1951–85 and aridization in the following years.

Fig. 5.
Fig. 5.

(a) Clustering results, (b) regionalization (clusters) of the AI during the decade 1951–2001, and (c) interannual changes in the AI with 11-yr moving-average smoothing curves. The colors and numbers in (b) and (a) correspond to the color lines and numbers in (c). Inset in (a) displays the link between the number of clusters and average within-cluster correlation. Inset in (b) shows the correlation coefficients between the center of each cluster and its members.

Citation: Journal of Applied Meteorology and Climatology 54, 10; 10.1175/JAMC-D-14-0207.1

Fig. 6.
Fig. 6.

(a) Location of weather stations with more than 40 yr of observations that display the month of the main peak summer rainfall (between May and October) in Mexico. (b) As in (a), but the symbols display the month when a second peak occurs.

Citation: Journal of Applied Meteorology and Climatology 54, 10; 10.1175/JAMC-D-14-0207.1

The second region was located southeast of the first region and included the semiarid and dry subhumid lands in the northern part of the Sierra Madre Oriental, the southern portion of the Mexican Altiplano, and along the trans-Mexican volcanic belt. The greatest peak in rainfall in this second region changed from August in the western portion of the Mexican Altiplano and the Pacific Ocean (Brito-Castillo et al. 2010) to September in the Gulf of Mexico watershed, east of the Sierra Madre Oriental (Fig. 6a). In some portions of the eastern highlands of the Mexican Altiplano, the main peak in rainfall occurred in May and June (Fig. 6a). This early season maximum across the Sierra Madre Oriental was attributed to a late penetration of deep cold air masses associated with abundant cloudiness, rainfall, and low temperatures (Cavazos 1999). Figure 6 shows the spatial distribution of the main peak (Fig. 6a) and second peak (Fig. 6b) of summer rainfalls (May–October) at stations displaying a midsummer drought (Magaña et al. 1999; i.e., a bimodal distribution of summer rainfall) in Mexico. The oscillations of AI in the second region were similar to those in the first region: a small rise in AI occurred in 1951–69, which gave way to slight aridization in the following years.

The periods of rises and declines in humidification across the third region, which was located in the center of the Pacific coast of Mexico, were alternating. The beginning of the study period was characterized by the slight rise in humidification. A rise was also observed in the period of the early 1980s to mid-1990s, while the transition between the two periods and in the end of the 1990s was marked with aridization. The amplitude of AI oscillation across the third region was higher than within the first and second regions. Warm-season rainfall distribution in the third region displayed a summer peak in July, with some stations irregularly dispersed along the highlands of the western Sierra Madre Occidental and northern portion of the trans-Mexican volcanic belt displaying August maximum (Fig. 6a). This late peak of rainfall might be associated with orographic controls of cloud and precipitation that added to rainfall variability over those areas (Giovannettone and Barros 2008).

The fourth region was located in the southeastern, wet humid part of Mexico along the coast of the Gulf of Mexico. The oscillation pattern of the aridity index was similar to changes in the index within the third region, that is, slow aridization of the area took place between the periods of rise in humidification during the first 20 years of the study period and after the mid-1990s. However, the amplitude of the oscillations and rate of changes in AI were larger than in the third region and all other regions. The main summer peak of rainfall in the fourth region was observed in September (Fig. 6a) with a secondary peak observed in July (Fig. 6b) and a decline in rainfall in July–August. This behavior in rainfall indicated that in fourth region a midsummer drought was also a climatic characteristic of seasonal rainfall variability, although the physical explanation of its occurrence should be different from that provided by Magaña et al. (1999). In this case, the main peak in rainfall observed in September was attributed to maximum activity of tropical cyclones (Jauregui 2003).

The fifth region included the Pacific coast area to the south of the Tropic of Cancer and southern Mexico with the exception of the Yucatan Peninsula. This region was predominantly subhumid and wet subhumid lands, with the inclusion of some dry subhumid and even semiarid lands near the Pacific coast of the Sierra Madre’s western slope. Weak, slow aridization of this area had been in progress up to 1979. Afterward, following stabilization of more humid conditions, a slight increase in humidification occurred until 1991. This region was subject to a peak in rainfall between June and September (Fig. 6a), with a second peak in August (Fig. 6b), with a decline in convective activity in late July–early August, causing a short period known as “canicula or midsummer drought” (Magaña et al. 1999), that reflected an anticyclone in southwestern Mexico north of the intertropical convergence zone (Giovannettone and Barros 2008).

The variation of the smoothing curve of the aridity index in the sixth region, across the Yucatan Peninsula, suggested that the slow rise in humidity took place at the beginning and at the second half of the study period. A decline occurred between the mid-1950s and 1970s. Similar to the fourth region, the peak in rainfall over the sixth region occurred in September (Fig. 6a), which was influenced by a late occurrence of tropical cyclones in the Atlantic Ocean.

As shown in Table 1, the SST temperature change in the El Niño region significantly affected the variability of the aridity index in the southern part of Mexico (regions 3, 4, 5, and 6) and less to the north, while the tropical Atlantic influenced the variability of the aridity index in the first region. The impact of sea surface temperature anomalies in the second region was apparently combined.

Table 1.

Correlations between the AI anomalies of each region and the anomalies of the indices of tropical Pacific and tropical Atlantic SST in the period 1982–2001.

Table 1.

To summarize, most of the land in the Mexican terrain (the first and second regions) had a slow aridization since the early 1980s. The decline in the aridity index in the early 1950s and late 1990s was caused by droughts in the area.

c. Relation between aridity index and AVHRR NDVI and their variability during the decade of 1992–2001 as compared with the decade of 1982–91

NDVI was considered an indicator of the range of the values of the aridity index. In Fig. 7, the contour lines of the index agreed well with the fields of annual means and seasonal means of NDVI. The linear relationship of aridity index and annual mean NDVI (the correlation coefficient 0.5–0.7 at the 0.05 significance level) was observed in regions with low vegetation on dry lands (Fig. 8). The largest coefficients of the linear correlation between the aridity index and NDVI in the humid zone did not exceed 0.56 on the Pacific coast of Oaxaca and decreased inland.

Fig. 7.
Fig. 7.

The AI (contours) and AVHRR NDVI values (colors) in 1982–2001: (a) annual mean; (b) May–September mean.

Citation: Journal of Applied Meteorology and Climatology 54, 10; 10.1175/JAMC-D-14-0207.1

Fig. 8.
Fig. 8.

Linear regression between AVHRR NDVI annual time series and average values of AI from 1982 to 2001. Points associated with low vegetation on dry lands are bordered by the red rectangle.

Citation: Journal of Applied Meteorology and Climatology 54, 10; 10.1175/JAMC-D-14-0207.1

The parameter of interest was the estimate of change in the aridity index and NDVI during the dry decade of 1992–2001, relative to the decade of 1982–91, which was not dry. Global warming was more intense during the dry decade of 1992–2001 (IPCC 2007). The drought at the end of the twentieth century mostly extended over the dry lands, where the aridity index declined (Fig. 9a). The most significant decrease in the index occurred on the western and eastern slopes of the Sierra Madre Occidental. The tendency for an increase in the index, with some isolated patches where it declined in central Mexico and Yucatan Peninsula, was observed in southern Mexico south of 20°N. It followed from Fig. 9 that, if we regarded NDVI as a characteristic of the aridity index pattern, the annual mean NDVI was less sensitive to droughts than mean seasonal NDVI. This was evident from comparing the area covered by mean annual NDVI (Fig. 9b) with the area covered by seasonal mean NDVI (Fig. 9c). On an annual scale, there was even a rise in NDVI during the dry decade in the northern part of the Sierra Madre Occidental, northeastern Mexico, and the Baja California Peninsula. Not surprisingly, the largest drop in NDVI occurred in the southern part of the Sierra Madre Occidental (state of Sinaloa) that is, the area with pronounced monsoon precipitation pattern and the northern part of the Sierra Madre Oriental. The area of decline in mean seasonal NDVI was significantly larger than the annual mean. The highlands of the Sierra Madre Occidental, which were dry subhumid and humid, did not have a moisture deficit during the dry decade in the end of the twentieth century, and the NDVI over these areas increased. This result was supported by the estimate of the monthly monsoon index (MMI) during the two above-mentioned decades over the highlands. The MMI increased in the dryer decade (1992–2001; Douglas et al. 1993). Note that the statistically significant decrease of AI had not resulted to have a significant change of the annual NDVI (with the exception of ​state of Sinaloa).

Fig. 9.
Fig. 9.

Changes in the AI and the AVHRR NDVI anomalies in 1992–2001 relative to 1982–91: (a) AI, (b) annual cycle of NDVI, and (c) NDVI (May–September). Areas with statistically significant differences are hatched.

Citation: Journal of Applied Meteorology and Climatology 54, 10; 10.1175/JAMC-D-14-0207.1

What caused the prolonged drought at the end of the twentieth century at the boundary of AI = 0.75) during the decades of 1982–91 and 1992–2001 (Figs. 10a,b). In 1982–91 this contour remained within the range of one standard deviation of the aridity index, which was an indication of stability of the index. During 1992–2001 the significant change in the boundary of dry lands occurred on the western slopes of the southern part of the Sierra Madre Occidental in southern Sinaloa. This was independently corroborated by the statistically significant decline in NDVI over these areas (Figs. 9b,c).

Fig. 10.
Fig. 10.

The 0.75 contour line of the AI during (a) 1982–91 and (b) 1992–2001. Areas with AI values lower than 1 std dev (based on 1951–80 data) areas are in green.

Citation: Journal of Applied Meteorology and Climatology 54, 10; 10.1175/JAMC-D-14-0207.1

d. EOF analysis of the aridity index

Spatial patterns of the first two EOF modes of the aridity index and the time series of its main components are shown in Fig. 11. The first mode of the EOF, which explained 28.2% of the total variance, had a main center in the mountainous region at the north of the Mexican Plateau (Fig. 11a). The second mode of the EOF, which explained 9.4% of the total variance, also had only one main center located in the Llanura Costera del Golfo region (Fig. 11b). We limited the analysis considering the first four EOF modes that together explained a slightly more than 50% of the total variability of AI (the third mode explained 8.6% and the fourth mode explained 6.2%, not shown). The centers of the first, third, and fourth mode were in the northern part of Mexico to the north of the Tropic of Cancer. Along with this, the centers of the third and fourth EOF modes were on the shores of the Gulf of California (not shown). The loading values of leading modes of AI were consistent with the loading values of the leading modes of the annual precipitation revealed in a previous study (Comrie and Glenn 1998) over the southwest United States and northern Mexico. The area covered by the leading modes in the Comrie and Glenn (1998) study corresponded well with the leading modes of AI in this work.

Fig. 11.
Fig. 11.

For (a) EOF 1 and (b) EOF 2, (left) the leading EOF modes of the annual AI in the period 1951–2001 and (right) their temporal variability (blue line) and 11-yr moving-average smoothing curves (red line). The percentages of explained variance for each mode are shown at the top of the right panels.

Citation: Journal of Applied Meteorology and Climatology 54, 10; 10.1175/JAMC-D-14-0207.1

The temporal oscillation of the first mode was characterized by a positive trend in the monsoon region in the first half of the period and a weak negative trend since the late 1970s. The high correlation between AI and precipitation made it possible to assert that the variability of AI with both EOF modes was strongly controlled by the North Atlantic subtropical anticyclone and its interaction with regional air masses and the continental elevated mixed layer. The trends of the smoothing curves of the temporal oscillation of EOF 1 and EOF 2 were opposite (Figs. 11a,b).

4. Discussion

In principle, the drought near the end of the twentieth century and the beginning of the twenty-first century can be regarded as part of the trend toward aridization predicted for subtropical areas in both hemispheres (IPCC 2007). The trend is accounted for a northward shift of the Hadley circulation cell; hence the northward propagation of dry subtropical conditions takes place (Lu et al. 2007). Also, climatic warming since the beginning of the twentieth century contributes to aridization of Mexican climates (Shein 2006).

The droughts of the second half of the twentieth century have a greater impact on the aridization of dry lands than humid lands because of the regional characteristics of the droughts. However, droughts in arid lands and anomalously wet conditions in the humid part of Mexico often occur simultaneously, which may be due to the interaction between easterly waves and the trade winds over the Gulf of Mexico and the Caribbean Sea (Méndez and Magaña 2010).

Of all droughts of the second half of the twentieth century, the emphasis of our study is on the two major droughts during the 1950s and 1990s. But we also show the relationship between the drought of the 1960s and the aridity index for the area, which includes the southern part of the Mexican Altiplano, the southern Sierra Madre, and the Yucatan Peninsula (Figs. 3 and 4b).

Our findings show that slow aridization (negative trend in the aridity index) is predominant during the second half of the twentieth century across the Mexican dry lands north of the Tropic of Cancer. Additionally, aridization begin in the 1970s for the second aridity index region, then in the mid-1980s for the first index region, and in the early 1990s for the third index region (see Fig. 5). The subhumid and wet humid lands along the southern part of the eastern Pacific (the fifth index region) undergo with very slow aridization since the 1950s. These results are largely long-term climatic oscillations inherent to tropical regions rather than the processes of anthropogenic warming (i.e., Seager 2007; Seager et al. 2010; Brito-Castillo 2012). Aridization in a large part of Mexico north of 21°N is enhanced at the end of the twentieth century by a long-term drought (Fig. 9).

5. Conclusions

Six regions in Mexico, with typical interannual changes in the aridity index, have been defined by the 1951–2001 meteorological dataset. Peak months of rainfall differ within the regions. In the first region, which includes the dry lands in northwestern Mexico and the northern part of the Mexican Altiplano, precipitation peaks in mid- and late summer, which is a distinctive feature of the monsoon circulation. Early season peak in rainfall is observed in some portions of the eastern highlands of the Mexican Altiplano, while late season peak in rainfall is observed in the plains of the Gulf of Mexico watershed and the Yucatan Peninsula in the second, fourth, and sixth regions. Along the Gulf of Mexico’s plains, and also in southern Mexico along the Pacific coast area to the south of the Tropic of Cancer (the fourth and fifth regions, respectively), a peak in rainfall between June and September is likely, with a second peak in August and a decline in convective activity in late July and early August, causing a short dry period known as midsummer drought. Moreover, along the Pacific coast in the third region, south of the trans-Mexican volcanic belt, the midsummer drought also is observed in meteorological stations. Apparently, the northern limit of the occurrence of the midsummer drought in the Pacific watershed in Mexico is the trans-Mexican volcanic belt.

Same-sign changes in the aridity index are typical over dry lands. Humidification slowly increases from the early 1950s through the mid-1970s. At the end of the century, this process changes to a slow aridization of dry lands. The prolonged droughts of the early 1950s and 1990s have little impact on the dry lands.

The largest interannual changes in the aridity index occur across the wet humid lands in southeastern Mexico, adjacent to the Gulf of Mexico. Of all regions, these are the most impacted by droughts during the early 1950s and 1980s. Subhumid central Mexico experiences slow aridization since the late 1970s. The lower aridity index from the early 1950s to the late 1990s is caused by droughts in this area.

Increased aridization of dry lands caused by long-term droughts during the last decade of the twentieth century does not result in a sizeable shift of the southern boundary of the dry lands. The only exception is the southern boundary (AI = 0.75) in the state of Sinaloa. In this area, the boundary moves southward and aridization intensifies.

The distinctive features of the aridization of Mexican dry lands are characterized by steady and extensive droughts, of which there are three (1948–57, 1960–65, and 1994–2003) in the second half of the twentieth century. The drought of the 1990s, which occurred during the warming by the end of the twentieth century, was smaller in area and less severe than the drought of the 1950s. It may be assumed that droughts of the 1990s were caused by the joint effect of sea surface temperature anomalies and the anthropogenic impact on the soil and vegetation. However, the drought of the 1990s cannot be easily linked to anthropogenic climate change. The results obtained here can be used in studies of possible anthropogenic impact on the drought of the twentieth century’s last decade in Mexico, which includes the changes in land use.

Acknowledgments

Special thanks are given to Ira Fogel of CIBNOR for editorial revisions. The authors appreciate the comments of the anonymous reviewers. This study was funded by the Universidad de Guadalajara (Grant PROINPEP 2012-2013), Consejo Nacional de Ciencia y Tecnología of Mexico (CONACYT Grant 83433-CB), REDESClim (Reg. 254533), and CIBNOR (Grants PC 1.0, PC 0.3, and 896-1).

REFERENCES

  • Alhamed, A., , S. Lakshmivarahan, , and D. J. Stensrud, 2002: Cluster analysis of multimodel ensemble data from SAMEX. Mon. Wea. Rev., 130, 226256, doi:10.1175/1520-0493(2002)130<0226:CAOMED>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Allen, R. G., , M. Smith, , L. S. Pereira, , and A. Perrier, 1994: An update for the calculation of reference evapotranspiration. Int. Comm. Irrig. Drain. Bull., 43, 3592.

    • Search Google Scholar
    • Export Citation
  • Allen, R. G., , L. S. Pereira, , D. Raes, , and M. Smith, 1998: Crop evapotranspiration: Guidelines for computing crop water requirements. FAO Irrigation and Drainage Paper 56, 300 pp. [Available online at http://www.fao.org/docrep/x0490e/x0490e00.htm.]

  • Brito-Castillo, L., 2012: Regional patterns of trends in long-term precipitation and stream flow observations: Singularities in a changing climate in Mexico. Greenhouse Gases: Emission, Measurement, and Management, G. Liu, Ed., In Tech, 387–412, doi:10.5772/32804.

  • Brito-Castillo, L., , E. R. Vivoni, , D. J. Gochis, , A. Filonov, , I. Tereshchenko, , and C. Monzon, 2010: An anomaly in the occurrence of the month of maximum precipitation distribution in northwest Mexico. J. Arid Environ., 74, 531539, doi:10.1016/j.jaridenv.2009.10.014.

    • Search Google Scholar
    • Export Citation
  • Cavazos, T., 1999: Large-scale circulation anomalies conducive to extreme precipitation events and derivation of daily rainfall in northeastern Mexico and southeastern Texas. J. Climate, 12, 15061523, doi:10.1175/1520-0442(1999)012<1506:LSCACT>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Comrie, A. C., , and E. C. Glenn, 1998: Principal components-based regionalization of precipitation regimes across the southwest United States and northern Mexico, with an application to monsoon precipitation variability. Climate Res., 10, 201215, doi:10.3354/cr010201.

    • Search Google Scholar
    • Export Citation
  • Cook, B. I., , R. Seager, , and R. Miller, 2011: Atmospheric circulation anomalies during two persistent North American droughts: 1932–1939 and 1948–1957. Climate Dyn., 36, 23392355, doi:10.1007/s00382-010-0807-1.

    • Search Google Scholar
    • Export Citation
  • Dai, A., , K. E. Trenberth, , and T. Qian, 2004: A global data set of Palmer drought severity index for 1870–2002: Relationship with soil moisture and effects of surface warming. J. Hydrometeor., 5, 11171130, doi:10.1175/JHM-386.1.

    • Search Google Scholar
    • Export Citation
  • Douglas, M. W., , R. A. Maddox, , K. Howard, , and S. Reyes, 1993: The Mexican monsoon. J. Climate, 6, 16651677, doi:10.1175/1520-0442(1993)006<1665:TMM>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Ekstrom, M., , P. D. Jones, , H. Fowler, , G. Lenderink, , T. A. Buishand, , and D. Conway, 2007: Regional climate model data used within the SWURVE project—1: projected changes in seasonal patterns and estimation of PET. Hydrol. Earth Syst. Sci., 11, 10691083, doi:10.5194/hess-11-1069-2007.

    • Search Google Scholar
    • Export Citation
  • Garcia, E., , R. Vidal, , and M. E. Hernandez, 1990: Las regiones climáticas de México (The climatic regions of Mexico). Atlas Nacional de México, Vol. 2, A. Garcia de Fuentes, Ed., Universidad Nacional Autónoma de México, Instituto de Geografia, 61 pp.

  • Giddings, L. S. M., , B. M. Rutherford, , and A. Maarouf, 2005: Standardized precipitation index zones for Mexico. Atmósfera, 18, 3356.

  • Giovannettone, J. P., , and A. P. Barros, 2008: A remote sensing survey of the role of the landform on the organization of orographic precipitation in central and southern Mexico. J. Hydrometeor., 9, 12671283, doi:10.1175/2008JHM947.1.

    • Search Google Scholar
    • Export Citation
  • Hare, F. K., 1993: Climate variations, drought and desertification. World Meteorological Organization 653, 45 pp.

  • Harris, I., , P. D. Jones, , T. J. Osborn, , and D. H. Lister, 2013: Updated high-resolution grids of monthly climatic observations: The CRU TS3.10 dataset. Int. J. Climatol., 34, 623–642, doi:10.1002/joc.3711.

  • Hulme, M., , R. March, , and P. D. Jones, 1992: Global changes in a humidity index between 1931–60 and 1961–90. Climate Res., 2, 122, doi:10.3354/cr002001.

    • Search Google Scholar
    • Export Citation
  • IPCC, 2007: Climate Change 2007: The Physical Science Basis. Cambridge University Press, 996 pp.

  • Jauregui, E., 2003: Climatology of land falling hurricanes and tropical storms in Mexico. Atmósfera, 16, 193204.

  • Kushnir, Y., , R. Seager, , M. Ting, , N. Naik, , and J. Nakamura, 2010: Mechanisms of tropical Atlantic SST influence on North American precipitation variability. J. Climate, 23, 56105628, doi:10.1175/2010JCLI3172.1.

    • Search Google Scholar
    • Export Citation
  • Lizárraga-Celaya, C., , C. J. Watts, , J. C. Rodriguez, , J. Garatuza-Payan, , R. L. Scott, , and J. Sáiz-Hernández, 2010: Spatio-temporal variation in surface characteristics over the North American monsoon region. J. Arid Environ., 74, 540548, doi:10.1016/j.jaridenv.2009.09.027.

    • Search Google Scholar
    • Export Citation
  • Lu, J., , G. A. Vecchi, , and T. Reichler, 2007: Expansion of the Hadley cell under global warming. Geophys. Res. Lett., 34, L06805, doi:10.1029/2006GL028443.

    • Search Google Scholar
    • Export Citation
  • Magaña, V., , J. Amador, , and S. Medina, 1999: The midsummer drought over Mexico and Central America. J. Climate, 12, 15771588, doi:10.1175/1520-0442(1999)012<1577:TMDOMA>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Méndez, M., , and V. Magaña, 2010: Regional aspects of prolonged meteorological droughts over Mexico and Central America. J. Climate, 23, 11751188, doi:10.1175/2009JCLI3080.1.

    • Search Google Scholar
    • Export Citation
  • Mendez-Barroso, L. A., , E. R. Vivoni, , Ch. J. Watts, , and J. C. Rodriguez, 2009: Seasonal and interannual relations between precipitation, surface soil moisture and vegetation dynamics in the North American monsoon region. J. Hydrol., 377, 5970, doi:10.1016/j.jhydrol.2009.08.009.

    • Search Google Scholar
    • Export Citation
  • Pavia, E. G., , and F. Graef, 2002: The recent rainfall climatology of the Mediterranean Californias. J. Climate, 15, 26972701, doi:10.1175/1520-0442(2002)015<2697:TRRCOT>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Penman, H. L., 1948: Natural evaporation from open water, bare soil and grass. Proc. Roy. Soc., 193A, 120145, doi:10.1098/rspa.1948.0037.

    • Search Google Scholar
    • Export Citation
  • Reich, R. M., , B. C. Aguirre-Bravo, , and V. A. Bravo, 2008: New approach for modeling climatic data with applications in modeling tree species distributions in the states of Jalisco and Colima, Mexico. J. Arid Environ., 72, 13431357, doi:10.1016/j.jaridenv.2008.02.004.

    • Search Google Scholar
    • Export Citation
  • Romesburg, C. H., 1984: Cluster Analysis for Researchers. Life Time Learning, 334 pp.

  • Scheff, J., , and D. M. W. Frierson, 2014: Scaling potential evapotranspiration with greenhouse warming. J. Climate, 27, 15391558, doi:10.1175/JCLI-D-13-00233.1.

    • Search Google Scholar
    • Export Citation
  • Seager, R., 2007: The turn of the century North American drought: Global context, dynamics, and past analogs. J. Climate, 20, 55275552, doi:10.1175/2007JCLI1529.1.

    • Search Google Scholar
    • Export Citation
  • Seager, R., , N. Naik, , M. Ting, , M. A. Cane, , N. Harnik, , and Y. Kushnir, 2010: Adjustment of the atmospheric circulation to tropical Pacific SST anomalies: Variability of transient eddy propagation in the Pacific–North America sector. Quart. J. Roy. Meteor. Soc., 136, 277296, doi:10.1002/qj.588.

    • Search Google Scholar
    • Export Citation
  • Shein, K. A., 2006: State of the climate in 2005. Bull. Amer. Meteor. Soc., 87 (6), S1–S102, doi:10.1175/BAMS-87-6-shein.

  • Stahle, D. W., and Coauthors, 2009: Early 21st-century drought in Mexico. Eos, Trans. Amer. Geophys. Union, 90, 8990, doi:10.1029/2009EO110001.

    • Search Google Scholar
    • Export Citation
  • Szekely, G. J., , and M. L. Rizzo, 2005: Hierarchical clustering via joint between-within distances: Extending Ward’s minimum variance method. J. Classif., 22, 151183, doi:10.1007/s00357-005-0012-9.

    • Search Google Scholar
    • Export Citation
  • Tereshchenko, I., , A. N. Zolotkrylin, , T. B. Titkova, , L. Brito-Castillo, , and C. O. Monzón, 2012: Seasonal variation of surface temperature-modulating factors in the Sonoran Desert in northwestern Mexico. J. Appl. Meteor. Climatol., 51, 15191530, doi:10.1175/JAMC-D-11-0160.1.

    • Search Google Scholar
    • Export Citation
  • Thornthwaite, C. W., 1948: An approach towards a rational classification of climate. Geogr. Rev., 38, 5594, doi:10.2307/210739.

  • Tian, D., , and C. Martinez, 2014: The GEFS-based daily reference evapotranspiration (ETo) forecast and its implication for water management in the southeastern United States. J. Hydrometeor., 15, 11521165, doi:10.1175/JHM-D-13-0119.1.

    • Search Google Scholar
    • Export Citation
  • Tucker, C. J., , J. E. Pinzon, , M. E. Brown, , D. Slayback, , E. W. Pak, , R. Mahoney, , E. Vermote, , and N. El Saleous, 2005: An extended AVHRR 8-km NDVI dataset compatible with MODIS and SPOT vegetation NDVI data. Int. J. Remote Sens., 26, 44854498, doi:10.1080/01431160500168686.

    • Search Google Scholar
    • Export Citation
  • UNCCD, 1994: United Nations convention to combat desertification in those countries experiencing serious drought and desertification, particularly in Africa. United Nations Treaty Series, Vol. 1954, 3. [Available online at https://treaties.un.org/pages/ViewDetails.aspx?src=TREATY&mtdsg_no=XXVII-10&chapter=27&lang=en.]

  • UNEP, 1992: World Atlas of Desertification. Edward Arnold, 69 pp.

  • UNESCO, 1979: Map of the world distribution of arid regions. UNESCO Man and the Biosphere (MAB) Tech. Note 7, 54 pp.

  • University of East Anglia Climatic Research Unit, 2013: CRU TS3.21: Climatic Research Unit (CRU) time series (TS) version 3.21 of high-resolution gridded data of month-by-month variation in climate (January 1901–December 2012). NCAS British Atmospheric Data Centre, accessed 2013–July 2014, doi:10.5285/D0E1585D-3417-485F-87AE-4FCECF10A992.

  • Vivoni, E. R., , H. A. Moreno, , G. Mascaro, , J. C. Rodriguez, , C. J. Watts, , J. Garatuza-Payan, , and R. L. Scott, 2008: Observed relation between evapotranspiration and soil moisture in the North American monsoon region. Geophys. Res. Lett., 35, L22403, doi:10.1029/2008GL036001.

    • Search Google Scholar
    • Export Citation
  • Watts, C. J., , R. L. Scott, , J. Garatuza-Payan, , J. C. Rodriguez, , J. H. Pruegger, , W. P. Kustas, , and M. Douglas, 2007: Changes in vegetation condition and surface fluxes during NAME 2004. J. Climate, 20, 18101820, doi:10.1175/JCLI4088.1.

    • Search Google Scholar
    • Export Citation
  • Wilks, D. S., 1995: Statistical Methods in the Atmospheric Sciences: An Introduction. Academic Press, 467 pp.

  • Zolotokrylin, A. N., 2003: Climatic Desertification (in Russian). Nauka, 246 pp.

Save