• Aceituno, P., 1988: On the functioning of the Southern Oscillation in the South American sector. Part I: Surface climate. Mon. Wea. Rev., 116, 505524, https://doi.org/10.1175/1520-0493(1988)116<0505:OTFOTS>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Aceituno, P., and A. Montecinos, 2000: Diurnal cycle over the South American Altiplano: Comparison of NCEP/NCAR reanalysis with upper-air observations during the Visviri field experiment. Preprints, Sixth Int. Conf. on Southern Hemisphere Meteorology and Oceanography, Santiago, Chile, Amer. Meteor. Soc., 412–413.

  • Aguirre, C., Ó. Pizarro, P. T. Strub, R. Garreaud, and J. A. Barth, 2012: Seasonal dynamics of the near-surface alongshore flow off central Chile. J. Geophys. Res., 117, C01006, https://doi.org/10.1029/2011JC007379.

    • Search Google Scholar
    • Export Citation
  • Barnett, T. P., J. C. Adam, and D. P. Lettenmaier, 2005: Potential impacts of a warming climate on water availability in snow-dominated regions. Nature, 438, 303309, https://doi.org/10.1038/nature04141.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Bourrel, L., and et al. , 2015: Low-frequency modulation and trend of the relationship between ENSO and precipitation along the northern to centre Peruvian Pacific coast. Hydrol. Processes, 29, 12521266, https://doi.org/10.1002/hyp.10247.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Bradley, R. S., M. Vuille, H. F. Diaz, and W. Vergara, 2006: Threats to water supplies in the tropical Andes. Science, 312, 17551756, https://doi.org/10.1126/science.1128087.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Chamorro Chávez, A., 2015: Stochastic and hydrological modelling for climate change prediction in the Lima region, Peru. Ph.D. dissertation, Universität Stuttgart, 119 pp., http://dx.doi.org/10.18419/opus-638.

    • Crossref
    • Export Citation
  • Dewitte, B., S. Illig, L. Renault, K. Goubanova, K. Takahashi, D. Gushchina, K. Mosquera, and S. Purca, 2011: Modes of covariability between sea surface temperature and wind stress intraseasonal anomalies along the coast of Peru from satellite observations (2000–2008). J. Geophys. Res., 116, C04028, https://doi.org/10.1029/2010JC006495.

    • Search Google Scholar
    • Export Citation
  • Drenkhan, F., M. Carey, C. Huggel, J. Seidel, and M. T. Oré, 2015: The changing water cycle: Climatic and socioeconomic drivers of water-related changes in the Andes of Peru. Wiley Interdiscip. Rev.: Water, 2, 715733, https://doi.org/10.1002/wat2.1105.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Eda, L. E. H., and W. Chen, 2010: Integrated water resources management in Peru. Procedia Environ. Sci., 2, 340348, https://doi.org/10.1016/j.proenv.2010.10.039.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Enfield, D. B., 1981: Thermally driven wind variability in the planetary boundary layer above Lima, Peru. J. Geophys. Res., 86, 20052016, https://doi.org/10.1029/JC086iC03p02005.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Garreaud, R. D., 1999: Multiscale analysis of the summertime precipitation over the central Andes. Mon. Wea. Rev., 127, 901921, https://doi.org/10.1175/1520-0493(1999)127<0901:MAOTSP>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Garreaud, R. D., and P. Aceituno, 2001: Interannual rainfall variability over the South American Altiplano. J. Climate, 14, 27792789, https://doi.org/10.1175/1520-0442(2001)014<2779:IRVOTS>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Garreaud, R. D., M. Vuille, and A. C. Clement, 2003: The climate of the Altiplano: Observed current conditions and mechanisms of past changes. Palaeogeogr. Palaeoclimatol. Palaeoecol., 194, 522, https://doi.org/10.1016/S0031-0182(03)00269-4.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Garreaud, R. D., M. Vuille, R. Compagnucci, and J. Marengo, 2009: Present-day South American climate. Palaeogeogr. Palaeoclimatol. Palaeoecol., 281, 180195, https://doi.org/10.1016/j.palaeo.2007.10.032.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Houston, J., and A. J. Hartley, 2003: The central Andean west-slope rainshadow and its potential contribution to the origin of hyper-aridity in the Atacama Desert. Int. J. Climatol., 23, 14531464, https://doi.org/10.1002/joc.938.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Illig, S., and et al. , 2014: Forcing mechanisms of intraseasonal SST variability off central Peru in 2000–2008. J. Geophys. Res. Oceans, 119, 35483573, https://doi.org/10.1002/2013JC009779.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Kain, J. S., 2004: The Kain–Fritsch convective parameterization: An update. J. Appl. Meteor., 43, 170181, https://doi.org/10.1175/1520-0450(2004)043<0170:TKCPAU>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Kalnay, E., and et al. , 1996: The NCEP/NCAR 40-Year Reanalysis Project. Bull. Amer. Meteor. Soc., 77, 437471, https://doi.org/10.1175/1520-0477(1996)077<0437:TNYRP>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Kaufman, L., and P. Rousseeuw, 1987: Clustering by means of medoids. Statistical Data Analysis Based on the L1 -Norm and Related Methods, Y. Dodge, Ed., North-Holland, 405–416.

  • Lavado Casimiro, W. S., J. Ronchail, D. Labat, J. C. Espinoza, and J. L. Guyot, 2012: Basin-scale analysis of rainfall and runoff in Peru (1969–2004): Pacific, Titicaca and Amazonas drainages. Hydrol. Sci. J., 57, 625642, https://doi.org/10.1080/02626667.2012.672985.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Lenters, J. D., and K. H. Cook, 1997: On the origin of the Bolivian high and related circulation features of the South American climate. J. Atmos. Sci., 54, 656678, https://doi.org/10.1175/1520-0469(1997)054<0656:OTOOTB>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Lenters, J. D., and K. H. Cook, 1999: Summertime precipitation variability over South America: Role of the large-scale circulation. Mon. Wea. Rev., 127, 409431, https://doi.org/10.1175/1520-0493(1999)127<0409:SPVOSA>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Lettau, H. H., 1967: Small to large-scale features of boundary layer structure over mountain slopes. Proc. Symp. on Mountain Meteorology, Vol. 26, Fort Collins, CO, Department of Atmospheric Science, Colorado State University, 1–73.

  • Lynch, B. D., 2012: Vulnerabilities, competition and rights in a context of climate change toward equitable water governance in Peru’s Rio Santa Valley. Global Environ. Change, 22, 364373, https://doi.org/10.1016/j.gloenvcha.2012.02.002.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Ma, H.-Y., C. R. Mechoso, H. Xiao, C.-M. Wu, Y. Xue, and F. De Sales, 2010: Connection between the South Pacific anti-cyclone, Peruvian stratocumulus, and the South American monsoon system. CLIVAR Exchanges, Vol. 15, International CLIVAR Project Office, Southampton, United Kingdom, 569–572.

  • McGlone, D., and M. Vuille, 2012: The associations between El Niño–Southern Oscillation and tropical South American climate in a regional climate model. J. Geophys. Res., 117, D06105, https://doi.org/10.1029/2011JD017066.

    • Search Google Scholar
    • Export Citation
  • Mlawer, E. J., S. J. Taubman, P. D. Brown, M. J. Iacono, and S. A. Clough, 1997: Radiative transfer for inhomogeneous atmospheres: RRTM, a validated correlated-k model for the longwave. J. Geophys. Res., 102, 16 66316 682, https://doi.org/10.1029/97JD00237.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Mohr, K. I., D. Slayback, and K. Yager, 2014: Characteristics of precipitation features and annual rainfall during the TRMM era in the central Andes. J. Climate, 27, 39824001, https://doi.org/10.1175/JCLI-D-13-00592.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Muñoz, R., 2008: Diurnal cycle of surface winds over the subtropical southeast Pacific. J. Geophys. Res., 113, D13107, https://doi.org/10.1029/2008JD009957.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Nakanishi, M., and H. Niino, 2006: An improved Mellor–Yamada level-3 model: Its numerical stability and application to a regional prediction of advection fog. Bound.-Layer Meteor., 119, 397407, https://doi.org/10.1007/s10546-005-9030-8.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Neukom, R., M. Rohrer, P. Calanca, N. Salzmann, C. Huggel, D. Acuña, D. A. Christie, and M. S. Morales, 2015: Facing unprecedented drying of the central Andes? Precipitation variability over the period AD 1000–2100. Environ. Res. Lett., 10, 084017, https://doi.org/10.1088/1748-9326/10/8/084017.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • OCHA, 2017: Peru: rainy season as of 10 May 2017. U.N. Office for the Coordination of Humanitarian Affairs Situation Rep. 10, 15 pp., https://reliefweb.int/sites/reliefweb.int/files/resources/redhum-pe-informe_de_situacion_10-eng-ocha-20170512.pdf.

  • Perry, L. B., A. Seimon, and G. M. Kelly, 2014: Precipitation delivery in the tropical high Andes of southern Peru: New findings and paleoclimatic implications. Int. J. Climatol., 34, 197215, https://doi.org/10.1002/joc.3679.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Perry, L. B., and et al. , 2017: Characteristics of precipitating storms in glacierized tropical Andean cordilleras of Peru and Bolivia. Ann. Amer. Assoc. Geogr., 107, 309322, https://doi.org/10.1080/24694452.2016.1260439.

    • Search Google Scholar
    • Export Citation
  • Rahn, D. A., and R. D. Garreaud, 2014: A synoptic climatology of the near-surface wind along the west coast of South America. Int. J. Climatol., 34, 780792, https://doi.org/10.1002/joc.3724.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Rasmussen, K. L., and R. A. Houze Jr., 2011: Orogenic convection in subtropical South America as seen by the TRMM satellite. Mon. Wea. Rev., 139, 23992420, https://doi.org/10.1175/MWR-D-10-05006.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Rau, P., L. Bourrel, D. Labat, P. Melo, B. Dewitte, F. Frappart, W. Lavado, and O. Felipe, 2017: Regionalization of rainfall over the Peruvian Pacific slope and coast. Int. J. Climatol., 37, 143158, https://doi.org/10.1002/joc.4693.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Rayner, N. A., D. E. Parker, E. B. Horton, C. K. Folland, L. V. Alexander, D. P. Rowell, E. C. Kent, and A. Kaplan, 2003: Global analysis of sea surface temperature, sea ice, and night marine air temperature since the late nineteenth century. J. Geophys. Res., 108, 4407, https://doi.org/10.1029/2002JD002670.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Rodwell, M. J., and B. J. Hoskins, 2001: Subtropical anticyclones and summer monsoons. J. Climate, 14, 31923211, https://doi.org/10.1175/1520-0442(2001)014<3192:SAASM>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Rutllant, J. A., H. Fuenzalida, and P. Aceituno, 2003: Climate dynamics along the arid northern coast of Chile: The 1997–1998 Dinámica del Clima de la Región de Antofagasta (DICLIMA) experiment. J. Geophys. Res., 108, 4538, https://doi.org/10.1029/2002JD003357.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Salzmann, N., C. Huggel, S. U. Nussbaumer, and G. Ziervogel, 2016: Setting the scene: Adapting to climate change—A large-scale challenge with local-scale impacts. Climate Change Adaptation Strategies—An Upstream–Downstream Perspective, N. Salzmann et al., Eds., Springer, 3–16.

    • Crossref
    • Export Citation
  • Schütze, M., 2015: Sustainable water and wastewater management in urban growth centres coping with climate change—Concepts for Lima Metropolitana (Perú). BMBF Final Rep., 55 pp., http://www.lima-water.de/en/documents.html?Menu=5.

  • Silva Dias, P. L., W. H. Schubert, and M. DeMaria, 1983: Large-scale response of the tropical atmosphere to transient convection. J. Atmos. Sci., 40, 26892707, https://doi.org/10.1175/1520-0469(1983)040<2689:LSROTT>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Skamarock, W. C., and J. B. Klemp, 2008: A time-split nonhydrostatic atmospheric model for weather research and forecasting applications. J. Comput. Phys., 227, 34653485, https://doi.org/10.1016/j.jcp.2007.01.037.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Takahashi, K., 2005: The annual cycle of heat content in the Peru Current region. J. Climate, 18, 49374954, https://doi.org/10.1175/JCLI3572.1.

  • Takahashi, K., 2012: Thermotidal and land-heating forcing of the diurnal cycle of oceanic surface winds in the eastern tropical Pacific. Geophys. Res. Lett., 39, L04805, https://doi.org/10.1029/2011GL050692.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Tewari, M., and et al. , 2004: Implementation and verification of the unified NOAH land surface model in the WRF model. 20th Conf. on Weather Analysis and Forecasting/16th Conf. on Numerical Weather Prediction, Seattle, WA, Amer. Meteor. Soc., 14.2A, https://ams.confex.com/ams/pdfpapers/69061.pdf.

  • Thompson, G., R. M. Rasmussen, and K. Manning, 2004: Explicit forecasts of winter precipitation using an improved bulk microphysics scheme. Part I: Description and sensitivity analysis. Mon. Wea. Rev., 132, 519542, https://doi.org/10.1175/1520-0493(2004)132<0519:EFOWPU>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Vera, C., and et al. , 2006: Toward a unified view of the American monsoon systems. J. Climate, 19, 49775000, https://doi.org/10.1175/JCLI3896.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Viviroli, D., H. H. Dürr, B. Messerli, M. Meybeck, and R. Weingartner, 2007: Mountains of the world, water towers for humanity: Typology, mapping, and global significance. Water Resour. Res., 43, W07447, https://doi.org/10.1029/2006WR005653.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Vuille, M., and F. Keimig, 2004: Interannual variability of summertime convective cloudiness and precipitation in the central Andes derived from ISCCP-B3 data. J. Climate, 17, 33343348, https://doi.org/10.1175/1520-0442(2004)017<3334:IVOSCC>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Vuille, M., R. S. Bradley, and F. Keimig, 2000a: Climate variability in the Andes of Ecuador and its relation to tropical Pacific and Atlantic sea surface temperature anomalies. J. Climate, 13, 25202535, https://doi.org/10.1175/1520-0442(2000)013<2520:CVITAO>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Vuille, M., R. S. Bradley, and F. Keimig, 2000b: Interannual climate variability in the central Andes and its relation to tropical Pacific and Atlantic forcing. J. Geophys. Res., 105, 12 44712 460, https://doi.org/10.1029/2000JD900134.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Vuille, M., G. Kaser, and I. Juen, 2008: Glacier mass balance variability in the Cordillera Blanca, Peru and its relationship with climate and the large-scale circulation. Global Planet. Change, 62, 1428, https://doi.org/10.1016/j.gloplacha.2007.11.003.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Wang, C., and P. C. Fiedler, 2006: ENSO variability and the eastern tropical Pacific: A review. Prog. Oceanogr., 69, 239266, https://doi.org/10.1016/j.pocean.2006.03.004.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • View in gallery

    (a) Model domain setup and study area (dark-gray area in D3), SW and NE corners of D1 are 23.48°S, 87.10°W and 14.34°N, 29.71°W; SW and NE corners of D2 are 21.17°S, 82.36°W and 4.96°S, 62.24°W; and SW and NE corners of D3 are 15.41°S, 79.79°W and 7.64°S, 69.72°W. (b) Terrain configuration of the study area (SRTM data), spatial distribution of annual precipitation amount of each rain gauge (black markers), and the subcatchments T1–T3 (gray rectangular line markers). The direct distance between Lima and the lowest-elevated station is around 65 km.

  • View in gallery

    Mean monthly precipitation (mm) for each rain gauge over the time period December 1992–November 2012. The gauging stations (left y axis) are ordered by descending elevation (right y axis).

  • View in gallery

    Precipitation against elevation at each rain gauge: (a) annual mean precipitation amount; (b) CV; and percentage of mean seasonal precipitation on mean annual precipitation in (c) DJF, (d) MAM, (e) JJA, and (f) SON.

  • View in gallery

    Leading mode of the geopotential height z at 200 hPa (m; shaded, scaled by 100) overlaid with mean seasonal z at 200 hPa over the time period 1992–2012 (black contours; m) for (a) DJF, (b) MAM, (c) JJA, and (d) SON; the red marker indicates Lima. (e) Corresponding time scores (gray bars indicate MAM).

  • View in gallery

    (a) Spatial distribution of the assignment of the rain gauges to the annual clustering based on k medoids and the mean annual precipitation amount (mm), (b) terrain height (m; color shaded) and aspect of each rain gauge based on SRTM data, (c) spatial distribution of the assignment of the rain gauges to the seasonal clustering based on k medoids, and (d) monthly precipitation amount for each medoid of the annual clustering for the time period 1992–2012 (mm; stacked bars; cluster 1–4 are labeled as CL1–CL4; see Table 4) overlaid with the time scores from the EOF analysis (see Fig. 4). The respective medoids in (a) and (c) are highlighted with a black border.

  • View in gallery

    Monthly frequencies of positive (gray) and negative (blue) precipitation gradients based on the medoids from cluster analysis overlaid with Niño-1+2 (solid red line) and Niño-3.4 (dashed red line) indices (SST anomalies; °C) along (a) T1, (b) T2, (c) T3, and (d) positive/negative gradients at all three transects.

  • View in gallery

    Wind field in uυ directions (vectors; m s−1) and wind speed (color shaded; m s−1) at 500 hPa for WRF simulations (dx = 4 km) at 1500 LST for (a) POS1, (b) POS2, (c) NEG1, and (d) NEG2. The study area is highlighted in the subdomain, the red dot marker shows the location of the rain gauge located at the highest elevation, and the diagonal line marker represents the cross section along T2.

  • View in gallery

    As in Fig. 7, but for column-integrated liquid water (purple contours; mm; interval = 0.25) and column-integrated precipitation hydrometeors (blue contours; mm; interval = 1.5).

  • View in gallery

    Vertical cross section of the equivalent-potential temperature θe (color shaded; K), the wind field in uw directions (vectors; m s−1), cloud water mixing ratio (blue contours; kg kg−1), and the total precipitation mixing ratio (red contours; kg kg−1) at 1500 LST for (a) POS1, (b) POS2, (c) NEG1, and (d) NEG2. The gray vertical marker line indicates the position of the uppermost rain gauge. The location of the cross section is displayed in Figs. 7 and 8.

  • View in gallery

    SST (color shaded; K) and wind field in uυ directions (vectors; m s−1) at the 850-hPa height level at 1500 LST for (a) POS1 and (b) NEG1; (c) difference of SST (shaded; K) between POS1 and NEG1. Also shown are vertical cross sections of wind speed (shaded; m s−1) overlaid with wind field in uw directions (vectors; m s−1) for (d) POS1 and (e) NEG1; the cross-sectional line is indicated in (c).

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 60 60 11
PDF Downloads 33 33 6

Cross-Scale Precipitation Variability in a Semiarid Catchment Area on the Western Slopes of the Central Andes

View More View Less
  • 1 Laboratory for Climatology and Remote Sensing, Philipps-University Marburg, Marburg, Germany
  • | 2 Institute for Modelling Hydraulic and Environmental Systems, Department of Hydrology and Geohydrology, University of Stuttgart, Stuttgart, Germany
  • | 3 Centro de Investigación Ambiental para el Desarrollo, Universidad Nacional Santiago Antúnez de Mayolo, Huaraz, Peru
  • | 4 Institute of Ecology, Technische Universität Berlin, Berlin, Germany
  • | 5 Laboratory for Climatology and Remote Sensing, Philipps-University Marburg, Marburg, Germany
© Get Permissions
Full access

Abstract

Spatiotemporal precipitation patterns were investigated on the western slopes of the central Andes Mountains by applying EOF and cluster analysis as well as the Weather Research and Forecasting (WRF) Model. In the semiarid catchment area in the highlands of Lima, Peru, the precipitation is assumed to be a cross-scale interplay of large-scale dynamics, varying sea surface temperatures (SSTs), and breeze-dominated slope flows. The EOF analysis was used to encompass and elucidate the upper-level circulation patterns dominating the transport of moisture. To delineate local precipitation regimes, a partitioning cluster analysis was carried out, which additionally should illustrate local effects such as the altitudinal gradient of the Andes. The results demonstrated that especially during the transition to the dry season, synoptic-scale circulation aloft controls the precipitation (correlation coefficients between 0.6 and 0.9), whereas in the remaining seasons the slope breezes due to the altitudinal gradient mainly determine the precipitation behavior. Further analysis with regard to the spatiotemporal precipitation variability revealed an inversion of the precipitation distribution along the elevational gradient within the study area, mainly during February (29%) and March (35%), that showed correlations with coastal SST patterns ranging between 0.56 and 0.67. WRF simulations of the underlying mechanisms disclosed that the large-scale circulation influences the thermally induced upslope flows while the strength of southeastern low-level winds related to the coastal SSTs caused a blocking of easterlies in the middle troposphere through a reduced anticyclonic effect. This interplay enables the generation of precipitation in the usually drier environment at lower elevations, which leads to a decrease in rainfall with increasing elevation.

© 2018 American Meteorological Society. For information regarding reuse of this content and general copyright information, consult the AMS Copyright Policy (www.ametsoc.org/PUBSReuseLicenses).

Corresponding author: Katja Trachte, katja.trachte@geo.uni-marburg.de

Abstract

Spatiotemporal precipitation patterns were investigated on the western slopes of the central Andes Mountains by applying EOF and cluster analysis as well as the Weather Research and Forecasting (WRF) Model. In the semiarid catchment area in the highlands of Lima, Peru, the precipitation is assumed to be a cross-scale interplay of large-scale dynamics, varying sea surface temperatures (SSTs), and breeze-dominated slope flows. The EOF analysis was used to encompass and elucidate the upper-level circulation patterns dominating the transport of moisture. To delineate local precipitation regimes, a partitioning cluster analysis was carried out, which additionally should illustrate local effects such as the altitudinal gradient of the Andes. The results demonstrated that especially during the transition to the dry season, synoptic-scale circulation aloft controls the precipitation (correlation coefficients between 0.6 and 0.9), whereas in the remaining seasons the slope breezes due to the altitudinal gradient mainly determine the precipitation behavior. Further analysis with regard to the spatiotemporal precipitation variability revealed an inversion of the precipitation distribution along the elevational gradient within the study area, mainly during February (29%) and March (35%), that showed correlations with coastal SST patterns ranging between 0.56 and 0.67. WRF simulations of the underlying mechanisms disclosed that the large-scale circulation influences the thermally induced upslope flows while the strength of southeastern low-level winds related to the coastal SSTs caused a blocking of easterlies in the middle troposphere through a reduced anticyclonic effect. This interplay enables the generation of precipitation in the usually drier environment at lower elevations, which leads to a decrease in rainfall with increasing elevation.

© 2018 American Meteorological Society. For information regarding reuse of this content and general copyright information, consult the AMS Copyright Policy (www.ametsoc.org/PUBSReuseLicenses).

Corresponding author: Katja Trachte, katja.trachte@geo.uni-marburg.de

1. Introduction

High mountain areas such as the Andes Mountains in South America play an important role in the water supply of adjacent lowland areas (Viviroli et al. 2007). Changes in climate in these areas can have a direct impact on water quality and quantity (Salzmann et al. 2016). Over recent decades, the central Andes have experienced a warming trend, but the development of precipitation shows no clear trend so far. There is evidence, however, that regions south of 11°S indicate a trend toward less precipitation and vice versa (Drenkhan et al. 2015). Nonetheless, knowledge of precipitation processes and precipitation distribution as a fundamental component of the hydrological cycle plays a major role in estimating water availability and in developing adaptation strategies. This is particularly true for semiarid and arid regions of the western slopes of the central Andes. In Peru, for example, the coastal areas along the Pacific Ocean have access to only 1.6% of the country’s total available water resources (Eda and Chen 2010). Cities located in this area, such as Lima, where major parts of the water supply originate from precipitation in the higher mountain areas, already suffer from water shortage today (Schütze 2015; Eda and Chen 2010). With respect to changes in climate, population and water demand, this situation is expected to be exacerbated (e.g., Neukom et al. 2015; Lynch 2012; Bradley et al. 2006; Barnett et al. 2005). On the other hand, this region is also exposed to extreme precipitation events. From January to May 2017, such events caused severe flooding and triggered landslides along the western coast of South America, which led to severe damage to the infrastructure and affected more than 1 million people in Peru (OCHA 2017).

Our study site is located at the Cordillera Occidental between −11.38° and −12.13°S (see Fig. 1), with the main focus being on the area northeast of Lima, Peru, between 1800 and 4300 m MSL. Lima is located in a desert region (−12°2′35″S, −77°1′41″W) and therefore strongly relies on the aforementioned water supply from these highlands. The aridity of the coastal areas is caused by the large-scale subsidence generating a quasi-permanent inversion in the lower troposphere (Rutllant et al. 2003). The mean annual precipitation in Lima is approximately 9 mm. This number increases to approximately 900 mm in the higher elevations as the influence of the inversion decreases, dividing the Pacific from the Atlantic watersheds (Chamorro Chávez 2015). The precipitation is generally induced through large-scale moisture transport from the east and diurnal wind systems generating an increase in the amount of rain along the slopes.

Fig. 1.
Fig. 1.

(a) Model domain setup and study area (dark-gray area in D3), SW and NE corners of D1 are 23.48°S, 87.10°W and 14.34°N, 29.71°W; SW and NE corners of D2 are 21.17°S, 82.36°W and 4.96°S, 62.24°W; and SW and NE corners of D3 are 15.41°S, 79.79°W and 7.64°S, 69.72°W. (b) Terrain configuration of the study area (SRTM data), spatial distribution of annual precipitation amount of each rain gauge (black markers), and the subcatchments T1–T3 (gray rectangular line markers). The direct distance between Lima and the lowest-elevated station is around 65 km.

Citation: Journal of Applied Meteorology and Climatology 57, 3; 10.1175/JAMC-D-17-0207.1

In general, the precipitation behavior in the central Andes is driven by large-scale atmospheric circulation systems providing the ambient environmental conditions, and thus follows a distinct seasonality. Most of the precipitation falls from October to April, with the main rainy season being from December to February (DJF). The upper-level anticyclone [Bolivian high (BH)] associated with upper-level divergence is well established and centered near 20°S and 60°W (Lenters and Cook 1997). In the lower troposphere, the intertropical convergence zone (ITCZ) is located in the Southern Hemisphere and the easterly trade winds transport warm moist air from the Atlantic Ocean over the Amazon basin to the high cordilleras of the central Andes Mountains. Together with the high insolation during the austral summer months, precipitation patterns are mainly driven by the diurnal cycle, with late afternoon convective rainfall that develops in the conditionally unstable environments. Furthermore, the height and length of the Andean Mountain chain strongly affect the precipitation distribution. By acting as a barrier to the trade winds, a zonal gradient is created with most rainfall occurring on the eastern slopes, while the western slopes experience a rain shadow effect. During austral winter [June–August (JJA)], when the ITCZ migrates northward, the central Andes experience their main dry season with little or no precipitation. The South Pacific high pressure system is well established and associated with strong subsidence (Rodwell and Hoskins 2001; Ma et al. 2010) suppressing convective activity. Additionally, an enhanced upper-level westerly flow inhibits the transport of moisture from the Amazon, which intensifies the rain shadow and low precipitation. Along the eastern Pacific coast, low-level winds often develop because of the Peru Current, which leads to upwelling of cold water (Aguirre et al. 2012; Rahn and Garreaud 2014) enhancing the convective inhibition. For more details on the South American monsoon system, see Aceituno (1988), Silva Dias et al. (1983), Garreaud (1999), and Vera et al. (2006), among others.

However, most studies in the central Andes are focused on the eastern slopes, and only a few report on the semiarid western flanks of this region. Perry et al. (2014) analyzed precipitation patterns in the Cordillera Vilcanota. The authors analyzed the diurnal cycle of heavy precipitation and used back-trajectory modeling for airmass history. Mohr et al. (2014) examined the organization of convective systems in the central Andes using precipitation gauge observations and remote sensing data with special focus on the eastern slopes. The authors revealed that precipitation is primarily of short duration and that events are mainly characterized by weak and shallow convection. Contrasts in precipitation occurrences between western and eastern slopes are mainly driven by the topography of the Andes. For the western and eastern slopes of Peru, Lavado Casimiro et al. (2012) highlight the differences in dry coastal and wet eastern areas on a basin scale, based on monthly precipitation data. The barrier effect of the terrain was also demonstrated by Houston and Hartley (2003), who reported on a rain shadow effect due to the Andes Mountains in a hyperarid region in the Atacama Desert. Another aspect that affects precipitation variability in this region is the ENSO phenomenon. Recently, Perry et al. (2017) analyzed spatiotemporal precipitation patterns in the Andes of Peru and Bolivia and highlighted the complex relationships with ENSO in these regions. The authors could further show that nighttime precipitation is mostly stratiform and is associated with an Amazonian moisture influx. A recent study related to precipitation variability along the western slopes of the Peruvian Andes was published by Rau et al. (2017). Here, the authors used a k-means clustering technique as a first step for regionalization of precipitation, which was followed by a regional vector method. Based on this approach, Rau et al. (2017) identified and characterized nine different precipitation regions based on monthly precipitation data and related these regions to SSTs in the Atlantic and Pacific.

The major objective of this study is to explore the spatiotemporal precipitation patterns in the catchment area of the highlands of Lima, taking into account both large-scale and small-scale influences. We hypothesize that the precipitation behavior in our study area along the western flanks of the Andes is organized by an interplay of large-scale dynamics, associated with varying sea surface temperatures (SSTs) and breeze-dominated slope flows generating positive and negative precipitation gradients along catchments. Daily precipitation measurements, atmospheric parameters derived from NCEP–NCAR reanalysis data (Kalnay et al. 1996) and Niño-1+2 and Niño-3.4 SSTs are used for the analysis. Further, the Weather Research and Forecasting (WRF) Model (Skamarock and Klemp 2008) is employed to examine the underlying dynamic processes affecting the variability in precipitation of the study site. With our analysis, we aim to assess (i) if the precipitation variability can be explained by the large-scale circulation as the main contributor to atmospheric environmental conditions inducing its formation, (ii) if different precipitation regimes in the study area can be delineated, and (iii) if the precipitation behavior is modified on the local scale by the altitudinal gradient of the Andes Mountains and varying coastal SSTs.

The next section gives a brief overview of the data used and the statistical and numerical methods encompassing the model setup. The results are described and discussed in the context of a regionalization of precipitation patterns as well as their driving features on the western slopes of the Andes. The findings are supported by numerical case studies.

2. Data and methods

For this study, the data from precipitation gauges in the catchment area northeast of Lima (see Table 1 and Fig. 1) are used to analyze precipitation variability. The data covers the general hydrological catchment area of the highlands of Lima and was provided as daily data by the Peruvian Meteorological and Hydrological National Service (SENAMHI). The time period considered encompasses December 1992–November 2012. From the available rain gauges, we selected 18 stations that are located above the inversion layer between 1800 and 4300 m MSL and have the most complete data records. Because of the high spatial and temporal heterogeneity of the rainfall in the study area, the stations show little mutual correlation. Therefore we limited quality control to removing obvious outliers.

Table 1.

Overview of the precipitation stations used for this study.

Table 1.

For the analysis of coastal SST patterns, which play a major role in the formation of precipitation (Vuille et al. 2000b; Garreaud et al. 2009), we used the Hadley Centre SST dataset (HadISST) data version 3 (Rayner et al. 2003).

To capture the main atmospheric circulation types as the main contributor for environmental conditions and to relate these types to the measured precipitation of the study region, an empirical orthogonal function (EOF) analysis was conducted. The application of EOF techniques in classifying atmospheric circulation modes as well as their relation to surface variables is shown by various studies (Vuille et al. 2000a; Vuille and Keimig 2004; McGlone and Vuille 2012). The EOF was applied to the geopotential height at 200 hPa (hereinafter z) derived from NCEP–NCAR reanalysis data (2.5° × 2.5°) on a seasonal basis for the considered time period (December 1992–November 2012). The suitability of the rather coarse reanalysis data in such complex areas was demonstrated by previous studies (e.g., Aceituno and Montecinos 2000; Vuille and Keimig 2004). To test the physical relationship with the observed precipitation variability, a Pearson correlation analysis on a monthly basis was performed.

For detection and delineation of the precipitation regimes in the heterogeneous study area, a partitioning cluster analysis based on the k-medoids algorithm (Kaufman and Rousseeuw 1987) was conducted, as also demonstrated by studies of Rau et al. (2017) and Lavado Casimiro et al. (2012). This clustering algorithm is related to k means but uses the data points as centers and is defined by
e1
where Ci is the ith cluster, ci is the medoid of the ith cluster, x is the respective data, and dist is the Euclidean distance. The aim of the cluster analysis is to divide the data into k partitions while minimizing the sum of pairwise dissimilarities (DIS) of the cluster medoids. A medoid represents a centrally located data point in the cluster, as the average dissimilarity to all objects is minimal. In our case, this means the aim is to group the gauging stations with comparable precipitation variability and thus detect different characteristics within the study area. The objects of the cluster analysis are the time series of the rain gauges (xi). This enables partitioning related to the temporal precipitation variability, while topographical influences such as elevation, slope, aspect, or exposure can be neglected. The objects are assigned to one cluster using a random partitioning. Each one reaches its final cluster when the similarity is greatest. The similarity is calculated based on the Euclidean distance [see Eq. (1)]. To account for the strong seasonality in the precipitation, we conducted the cluster analysis on the entire time period (December 1992–November 2012) as well as on each season, that is, DJF, March–April (MAM), JJA, and September–November (SON). The appropriate numbers of clusters are determined by means of the isolation of each cluster.

Based on the medoids (representative rain gauge of the respective cluster) of the annual clustering, precipitation gradients along the three subcatchments (Fig. 1b) are emphasized by calculating the difference between the highest and lowest located medoid with respect to the elevation. The purpose of this analysis is to account for the strong altitudinal gradient. To consider their persistence, the occurrences of precipitation gradients are calculated on the basis of 3-day means.

To reflect the precipitation behavior and further investigate the dynamic coupling between the altitudinal gradient, the atmosphere and the SSTs, the WRF Model was used for specific precipitation events in the study area. WRF is a fully compressible, nonhydrostatic gridbox model with a terrain-following vertical coordinate. The model was applied in a 3-nested domain setup (Fig. 1a) in a two-way nesting mode: the outer domain has 155 grid points in the east–west and 105 grid points in the north–south directions with a grid size of 36 km to capture the main driving features of South America affecting the study site. Domain 2 has 163 by 136 grid points (12 km). Domain 3 focuses on the catchment area of Lima with 244 by 193 grid points and a spatial resolution of 4 km. The simulations are driven by the NCEP–NCAR reanalysis (Kalnay et al. 1996), and the surface properties (land coverage and topography) are described by the 24 land categories and the GTOPO30 terrain data of the USGS. Each simulation was carried out for 24 h with a spinup of 48 h. The parameterization schemes used are summarized in Table 2. We used Thompson scheme since it is suitable for high-resolution simulations. For the planetary boundary layer (PBL) we selected a local closure scheme that predicts subgrid turbulent kinetic energy.

Table 2.

WRF parameterization schemes considered for the case studies and respective references.

Table 2.

3. Results

a. General precipitation variability

First, the general precipitation variability in the catchment area is analyzed with respect to the altitudinal gradient in order to assess its influence. The mean annual precipitation in the study area ranges between 70 and 910 mm over the time period December 1992 to November 2012 (Fig. 1) and shows a distinct positive height gradient, that is, the rain gauges located at higher altitudes receive more precipitation than the stations farther downslope. The annual cycle of monthly mean precipitation at each station (ordered by elevation) illustrates similar variations between the rain gauges and deviations only occur in terms of the precipitation amount (Fig. 2). However, a shift in the peak during February and March can be observed that indicates different influences on the precipitation formation. Furthermore, the onset of the rainy season is shifted from August to November with decreasing elevation.

Fig. 2.
Fig. 2.

Mean monthly precipitation (mm) for each rain gauge over the time period December 1992–November 2012. The gauging stations (left y axis) are ordered by descending elevation (right y axis).

Citation: Journal of Applied Meteorology and Climatology 57, 3; 10.1175/JAMC-D-17-0207.1

This behavior is reflected in the mean annual precipitation (Fig. 3a) as well as in the coefficients of variation (CV; ratio of the standard deviation to the mean; Fig. 3b) against the elevation of each gauging station. Both indicate a linear relationship with elevation owing to a high confidence with an explained variance of 0.81 and 0.76 (p value <0.05), respectively. The CV has a negative direction, which means that the highest variability occurs at the gauging stations located at the lowest altitudes.

Fig. 3.
Fig. 3.

Precipitation against elevation at each rain gauge: (a) annual mean precipitation amount; (b) CV; and percentage of mean seasonal precipitation on mean annual precipitation in (c) DJF, (d) MAM, (e) JJA, and (f) SON.

Citation: Journal of Applied Meteorology and Climatology 57, 3; 10.1175/JAMC-D-17-0207.1

The percentage of mean seasonal precipitation on mean annual precipitation for each station (Figs. 3c–f) shows that most of the rainfall events occur during the austral summer (50%–60%; Fig. 3c), as also reported in previous studies (e.g., Vuille et al. 2000b; Garreaud et al. 2003; Perry et al. 2014). In contrast, the amount decreases to 0%–4% during JJA (Fig. 3d). This season represents the main dry period, where the gauging stations located at lower elevations recorded no precipitation at all because of the influence of the inversion layer suppressing any convective activity.

The linear relationship between elevation and the percentage of mean seasonal precipitation on mean annual precipitation is obvious, though differences during the seasons can be observed. For the rainy seasons (DJF, MAM) a negative relationship exists, while the drier seasons (JJA, SON) show the opposite behavior. Moreover, the strength of the linear relationship clearly varies between each season, but this is not solely due to the percentage of rain amount. During DJF, a significant negative correlation occurred with an explained variance of 0.87, indicating a strong impact of the altitudinal gradient during these months. Higher located stations are more affected by intrusions of moist air from the Amazon caused by stronger easterlies. For MAM, a clear decrease in the explained variance (0.37) appeared, which is caused by explicit deviations from the regression line of the rain gauges located at lower elevations. Even during JJA and SON, this linear relationship has a higher confidence/greater strength with values of 0.54 and 0.74, respectively. The reason for the low explained variance in MAM, and thus the decrease in the altitudinal forcing of the precipitation variability, might be the varying seasonal influences in the atmospheric environmental conditions (BH, South Pacific high) and will be analyzed in the next sections.

b. Large-scale dynamics

To extract the seasonal signal of large-scale circulation patterns, we conducted an EOF analysis based on z at 200 hPa. Figure 4 presents the spatial modes and corresponding temporal evolution of the leading EOF, which explains 44% (DJF), 73% (MAM), 33% (JJA), and 56% (SON) of the respective total variances. The leading EOF was the sole focus and is associated with the precipitation variability as discussed below, while the other modes revealed no physical relationships. This is also true for the second EOF in DJF, although it explains 30% of the total variances.

Fig. 4.
Fig. 4.

Leading mode of the geopotential height z at 200 hPa (m; shaded, scaled by 100) overlaid with mean seasonal z at 200 hPa over the time period 1992–2012 (black contours; m) for (a) DJF, (b) MAM, (c) JJA, and (d) SON; the red marker indicates Lima. (e) Corresponding time scores (gray bars indicate MAM).

Citation: Journal of Applied Meteorology and Climatology 57, 3; 10.1175/JAMC-D-17-0207.1

In DJF, a monopole pattern with an overall positive loading with the strongest anomalies in the z field (>0.08) in the northern parts can be observed. This pattern represents the dominance of the BH during this season generated by heating effects over the South American continent as described by Lenters and Cook (1997). The austral winter (JJA) features a dipole pattern with a dipole axis in the southwest–northeast direction. The mode describes the weakened BH associated with northward-migrating westerlies from the high latitudes. Interestingly, only in MAM is a dominant zonal structure in the upper-level circulation featured, leading to predominately zonal winds, while the remaining seasons show a meridional component. During the transition to the dry season the mode portrays the weakening of the BH and the South American monsoon system related to a rather zonal flow regime.

Table 3 presents the correlation coefficients, which determine whether the precipitation variability of each station is related to the variability of the leading mode. Generally, a relationship between the defined anomalous patterns of z and precipitation occurred. However, differences regarding the season can be recognized. During DJF, the rain gauges show a relationship to z but with a weak performance (correlation coefficients between 0.1 and 0.5). The situation changes during MAM, when strong correlation coefficients ranging between 0.6 and 0.9 appear. Comparable results are obtained in SON, but with a negative sign (negative loadings) and a rather moderate signal. The temporal consistency between precipitation and the loadings of the first EOF indicates that during the transition from wet to dry season fluctuations in the upper-level circulation play a major role in controlling the precipitation variability. Zonal structures dominate, particularly in MAM, because of the northward migration of the ITCZ and the western winds aloft from the midlatitudes (Fig. 4b). This reduced meridional baroclinicity seems to be the crucial structure in the upper-level atmosphere when driving the precipitation variability consistent with Garreaud and Aceituno (2001) and Vuille and Keimig (2004).

Table 3.

Cross-correlation coefficients of the leading modes and the precipitation amount (mm) at each rain gauge. (Boldface values indicate significance with p values <0.05.)

Table 3.

c. Local precipitation regimes

Next, a partitioning cluster analysis is used to delineate different rain regimes and analyze the varying forcing on the precipitation variability. As described, the clustering was performed on the time series of the rain gauges to delineate different precipitation regimes in the entire time period (annual basis) as well as on a seasonal basis. In addition to the temporal component, its distribution/location in the study site additionally allows for a spatial characterization by highlighting the reflectance of topographical influences in the clustering (Figs. 5a,b). The results of the cluster analysis are summarized in Table 4.

Fig. 5.
Fig. 5.

(a) Spatial distribution of the assignment of the rain gauges to the annual clustering based on k medoids and the mean annual precipitation amount (mm), (b) terrain height (m; color shaded) and aspect of each rain gauge based on SRTM data, (c) spatial distribution of the assignment of the rain gauges to the seasonal clustering based on k medoids, and (d) monthly precipitation amount for each medoid of the annual clustering for the time period 1992–2012 (mm; stacked bars; cluster 1–4 are labeled as CL1–CL4; see Table 4) overlaid with the time scores from the EOF analysis (see Fig. 4). The respective medoids in (a) and (c) are highlighted with a black border.

Citation: Journal of Applied Meteorology and Climatology 57, 3; 10.1175/JAMC-D-17-0207.1

Table 4.

Cluster analysis based on k means and representative gauging station (medoid) for each considered time period.

Table 4.

On an annual basis (Year), four clusters related to its temporal precipitation variability can be differentiated. The highest situated rain gauge (Milloc at 4361 m MSL; see Table 3) is isolated in a single cluster, and the remaining stations are more or less equally assigned to the remaining clusters. Regarding the isolation, cluster 3 shows the best performance, while cluster 1 describes the most inhomogeneous class. Overall, the height dependency of precipitation is reflected in the spatial distribution of the clusters by means of a partitioning in terms of the altitude of the respective rain gauge. However, an additional north–south gradient can be recognized. The seasonal cluster results display varying influences, likely because of the strong seasonality of the upper-level large-scale circulation as described by Lenters and Cook (1997) and Garreaud (1999) (see Figs. 24). During the austral summer, the rain gauges can be grouped into five clusters representing both a zonal and meridional differentiation of precipitation regimes. Additionally, topographical effects can be noticed when stations with opposing aspects (see Table 1 and Fig. 5b) are grouped into different clusters. This also applies for MAM. The remaining seasons are only clustered into three groups. The assignment to the clusters changes with respect to the season, which again points to varying influences of the upper-level circulation providing the atmospheric environmental conditions for precipitation formation also observed in the correlation analysis (Table 3). In particular during MAM, the meridional partitioning is more strongly pronounced than the zonal partitioning. As mentioned in the previous section, this season represents the transition to dry conditions accompanied by a weakening of the BH and a northward shift of the ITCZ. In the dry season (JJA), all stations are grouped together, except for the two uppermost rain gauges. This means that nearly all stations are under the same environmental influence: likely the enhanced upper-level westward airflow, which weakens the transport of moist air from the Amazon and the dominating South Pacific high pressure system with its characterizing inversion layer suppressing convective activity (Garreaud 1999; Perry et al. 2014). In contrast, the rain gauges located at higher altitudes are affected by intrusions of moist air from the Amazon, which leads to the different precipitation regimes. In SON, a comparable clustering can be registered, but with slightly more gauging stations in cluster 2 (Table 4).

The monthly precipitation amount of each medoid of the annual clustering (here CL1–CL4; see Table 4 for rain gauge) overlaid with the time scores of the EOF analysis is illustrated in Fig. 5d. It is obvious that the objects are clustered with respect to the precipitation amount, but differences also occur in terms of the interannual variability. Moreover, this result highlights both the physical relationship to the large-scale circulation patterns and the altitudinal gradient as the main driver for the precipitation variability in the study area.

d. Precipitation gradients

Because the altitudinal gradient of the Andes Mountains plays a major role in the precipitation distribution in our study area, we further examine the precipitation occurrence along the three subcatchments (see Fig. 1). The resulting transects are summarized in Table 5. Here, not only positive precipitation gradients but also negative gradients are detected. The latter mostly occur during February (22%–25%) and March (22%–37%), while positive gradients are spread evenly across the rainy months. The absolute frequencies of occurrences show that the negative precipitation gradients show strong local formations with the highest frequencies along T1. In contrast the positive precipitation gradients are more equally distributed along the transects.

Table 5.

Percentage of positive and negative gradients (with the absolute frequencies over the entire time period in parentheses) per month for each transect (T1–T3) and simultaneously occurring at all transects. Correlation coefficients between absolute frequencies of gradients and respective Niño regions SST values are given in the bottom two rows. Boldface values indicate significance with p values <0.05.

Table 5.

Figure 6 presents the monthly frequencies of both precipitation gradients for each transect as well as for the entire region (positive–negative gradient at each transect) over the considered time period. The temporal distribution illustrates that peaks in the frequencies of negative gradients with a corresponding decrease in the frequencies of positive gradients are subject to both a seasonal and an interannual variability. In particular, during the strong ENSO warm-phase event in 1997/98, an increase in the frequencies of negative precipitation gradients in the study area emerges.

Fig. 6.
Fig. 6.

Monthly frequencies of positive (gray) and negative (blue) precipitation gradients based on the medoids from cluster analysis overlaid with Niño-1+2 (solid red line) and Niño-3.4 (dashed red line) indices (SST anomalies; °C) along (a) T1, (b) T2, (c) T3, and (d) positive/negative gradients at all three transects.

Citation: Journal of Applied Meteorology and Climatology 57, 3; 10.1175/JAMC-D-17-0207.1

Along the coastal areas of Peru, precipitation increases during El Niño events (Vuille et al. 2000b; Vera et al. 2006; Garreaud et al. 2009; Bourrel et al. 2015), so we tested statistical relations between SST and precipitation using two ENSO indices, that is, Niño-1+2 and Niño-3.4 regions SST anomalies and values. The Niño-1+2 and Niño-3.4 SST anomalies showed no relationship regardless of the frequencies of occurrences of positive and negative precipitation gradients (not shown in Table 5). In contrast, the Niño-1+2 regions SST values exhibit a signal with correlation coefficients with a maximum of 0.67 for the negative precipitation gradients, particularly along T1 and T3. For the positive cases, the signal is reversed, with the strongest consistency between Niño-1+2 SST values and the occurrences of positive precipitation gradients along T3 (0.37).

e. Underlying dynamic mechanisms of precipitation gradients

To investigate the underlying dynamic mechanisms of interacting large-scale circulation patterns, altitudinal gradients and the characteristics of SSTs leading to the detected precipitation gradients, the WRF Model was applied. For this reason, four case studies (Table 6) representing conditions of two positive (POS1, POS2) and two negative (NEG1, NEG2) precipitation gradient events were explored. For the negative cases, a nonanomalous (NEG1) and an anomalous (NEG2) atmospheric situation are differentiated to account for the ENSO phenomenon.

Table 6.

Four selected case studies and the respective daily rain amounts (mm) along the subcatchments observed at each medoid derived from the annual clustering (see Figs. 1 and 5a).

Table 6.

1) Wind field

Figure 7 illustrates the horizontal wind field (vectors; m s−1) and wind speed (shaded; m s−1) at the 500-hPa height level of the wider area of the study region at 1500 LST. The 500-hPa level instead of the 200-hPa level was selected to investigate the wind field close to the Andes crest. For POS1 the wind velocities in the study site are calm and the highest values (6 m s−1) evolve in the Amazon basin with a southerly wind direction. The second positive precipitation gradient shows a clear northeastern wind direction, and the highest velocities occur near the crest of the western flanks of the Andes Mountains (5–6 m s−1), while velocities decrease toward the Pacific Ocean (3–4 m s−1). In contrast, NEG1 demonstrates overall stronger wind speeds, with up to 8 m s−1 in the upper Andes and 5 m s−1 near the Pacific coast. The wind direction shows a dominant easterly component. Interestingly, NEG2 is more similar to POS2 in both wind speed and direction, although it describes a negative precipitation gradient. The former typically decreases during ENSO warm phases.

Fig. 7.
Fig. 7.

Wind field in uυ directions (vectors; m s−1) and wind speed (color shaded; m s−1) at 500 hPa for WRF simulations (dx = 4 km) at 1500 LST for (a) POS1, (b) POS2, (c) NEG1, and (d) NEG2. The study area is highlighted in the subdomain, the red dot marker shows the location of the rain gauge located at the highest elevation, and the diagonal line marker represents the cross section along T2.

Citation: Journal of Applied Meteorology and Climatology 57, 3; 10.1175/JAMC-D-17-0207.1

2) Clouds/precipitation patterns

Next, the occurrence of cloud and precipitation patterns are analyzed in the study area generated by the WRF Model based on the column-integrated liquid water (indicating clouds) and column-integrated precipitation hydrometeors for the same time step. As presented in Fig. 8, typical afternoon cloud/precipitation patterns develop on the western slopes above the inversion layer, presumably because of diabatic heating effects and subsequently upslope flows that trigger convective cells over the crest (Rasmussen and Houze 2011).

Fig. 8.
Fig. 8.

As in Fig. 7, but for column-integrated liquid water (purple contours; mm; interval = 0.25) and column-integrated precipitation hydrometeors (blue contours; mm; interval = 1.5).

Citation: Journal of Applied Meteorology and Climatology 57, 3; 10.1175/JAMC-D-17-0207.1

In POS1 (Fig. 8a), cloud and precipitation patterns can be identified in the study site with the typical formation of smaller cells. Specifically, along the western slopes, clouds are produced, while at the crest, precipitation also occurs, which develops the assumed positive gradient. For POS2 (Fig. 8b) the positive gradient is more featured because cloud/precipitation patterns are created only in the higher-elevated regions, presumably because of shallow or moderate convection. The situation is modified in NEG1 and NEG2, where most of the cloud/precipitation fields evolve in the lower-elevation areas of the western slopes. This is particularly true for NEG1, where in the vicinity of the highest rain gauge (red marker in Fig. 8), neither cloud nor precipitation fields occur. In NEG2 the cloud/precipitation patterns are spread rather equally, and yet there is a larger amount of hydrometeors in the lower altitudes, which denotes the negative gradient. Overall, the WRF Model generates cloud and precipitation fields that are consistent with the observed rain distribution along the considered transects (Table 6).

3) Atmospheric environmental conditions and slope breezes

To take a closer look at the local-scale dynamics as well as the environmental conditions of the atmosphere during the four selected situations, a vertical cross section through the cloud/precipitation patterns along catchment T2 is investigated (Fig. 9). This cross section is presented in Figs. 7 and 8.

Fig. 9.
Fig. 9.

Vertical cross section of the equivalent-potential temperature θe (color shaded; K), the wind field in uw directions (vectors; m s−1), cloud water mixing ratio (blue contours; kg kg−1), and the total precipitation mixing ratio (red contours; kg kg−1) at 1500 LST for (a) POS1, (b) POS2, (c) NEG1, and (d) NEG2. The gray vertical marker line indicates the position of the uppermost rain gauge. The location of the cross section is displayed in Figs. 7 and 8.

Citation: Journal of Applied Meteorology and Climatology 57, 3; 10.1175/JAMC-D-17-0207.1

The environmental conditions are examined in terms of the equivalent potential temperature θe as an indicator of the energy content of the air. Its values are quasi-constant within an air mass. The quantity θe is defined by
e2
where θ is the potential temperature, Le is the latent heat of evaporation, cp represents the specific heat content, TLCL is the temperature at the lifting condensation level (LCL), and wυ is the mixing ratio for water vapor.

As expected near the Pacific coast (Fig. 9), an inversion layer exists indicated by e/dz > 0, which is strongest for POS2 (θe ~ 300 K) and weakest during the strong El Niño event (NEG2 with θe = 338 K). On the western slopes following the diurnal cycle, local circulation systems such as anabatic winds develop in each case, as also reported by Rutllant et al. (2003). However, differences in the strength and the resulting cloud formations can be recognized. For the two positive cases (Figs. 9a,b), the thermally driven flows cause clouds and precipitation (solid red and blue lines) in an unstable environment (e/dz < 0) at the top of the slope. While deep convection was initiated in POS1, POS2 represents moderate convective activity, as mentioned above. Along the slopes, shallow clouds and precipitation evolve because of adiabatic cooling during the upslope motion. The result is the usually dominating positive precipitation gradient featured on both monthly and annual bases (see Figs. 2 and 3a). This situation is clearly modified in NEG1, where most of the rain and clouds appear at the flanks rather than near the crest of the western slopes. Here, we can observe that the unstable atmospheric stratification accompanied with cloud and precipitation formation is developed farther downslope, while in the vicinity of the uppermost located rain gauge (gray marker line in Fig. 9), convective activity is not induced. For NEG2, an overall warmer and moister atmosphere with a noticeable weaker inversion layer in the coastal areas can be recognized. As already assumed in Fig. 8, precipitation occurs at both height levels but with a larger amount at lower altitudes. Considering the wind field in the upper troposphere between 300 and 200 hPa, further differences can be observed. With the exception of NEG1, an easterly upper airflow is generated, which is explicitly developed throughout the atmosphere in POS2. In contrast, NEG1 presents rather weak velocities, and between the 400- and 300-hPa levels above the western flank, even westerlies are produced. Additionally, in this case, stronger winds near the surface in the highlands can be registered. NEG2 shows both a weaker upper-level as well as lower-level flow regime with an easterly direction.

4) Coastal SST patterns and winds

In our study area the SST plays a major role in the precipitation variability (Vuille et al. 2000a,b; Wang and Fiedler 2006) and this was also detected in our analysis (Table 5). Moreover, as described in Aguirre et al. (2012), there is a connection between SST patterns and low-tropospheric coastal winds affecting diabatic heating effects and vertical mixing. To study the impact of the SST patterns on the precipitation variability, Figs. 10a,b show the wind field at the 850-hPa height level associated with WRF SST for POS1 and NEG1 only. In both cases, a lower-tropospheric northward flow regime with comparable strength can be recognized, although discrepancies appear. While in POS1, the jet describes the typical direction along the coast, NEG1 exhibits an onshore component likely because of sea-breeze developments in the diurnal cycle. Further differences became evident considering the SSTs. Generally, in the vicinity of the coastal areas, cold SSTs occur (288 K), representing the upwelling and the Peru Current, while with increasing distance from the Peruvian coast, the SST increases as well (Takahashi 2005; Illig et al. 2014). When we compare POS1 and NEG1, discrepancies in this general pattern can be detected. For the latter, the area of colder SSTs is larger with an expansion along the coast, whereas for the same time POS1 obtained overall warmer SSTs (300 K southwest of the study site). By subtracting both SSTs (Fig. 10c), the assumed differences are highlighted and quantified. In particular, near the coast, positive deviations of up to 3.5 K arise, indicating colder values in NEG1. On the other hand, NEG1 reveals higher SSTs in the northwestern part of the domain.

Fig. 10.
Fig. 10.

SST (color shaded; K) and wind field in uυ directions (vectors; m s−1) at the 850-hPa height level at 1500 LST for (a) POS1 and (b) NEG1; (c) difference of SST (shaded; K) between POS1 and NEG1. Also shown are vertical cross sections of wind speed (shaded; m s−1) overlaid with wind field in uw directions (vectors; m s−1) for (d) POS1 and (e) NEG1; the cross-sectional line is indicated in (c).

Citation: Journal of Applied Meteorology and Climatology 57, 3; 10.1175/JAMC-D-17-0207.1

A vertical cross section (see Figs. 10d,e) illustrates the near-coastal flow regime throughout the atmosphere. Based on the wind speed, it is apparent that in POS1 (Fig. 10d), an offshore lower-level wind is developed over the Pacific Ocean that is lacking in NEG1 (Fig. 10e). Only near the Peruvian coast does an onshore flow appear as a result of a diurnal sea-breeze system. In contrast, NEG1 reveals an overall flow from southwestern directions. That means the developments of the low-tropospheric coastal flow and the sea breezes are strongly related to the characteristics of the SST patterns off the Peruvian coast, as also reported by Dewitte et al. (2011). In the middle and upper troposphere, further differences become evident. While POS1 displays weak westerlies (near 400 hPa) and typical descending air masses over the Pacific Ocean, NEG1 illustrates a layer (700–500 hPa) of strong easterlies already visible in Fig. 7c that blocks the large-scale subsidence.

4. Discussion and conclusions

As hypothesized, interplay between the upper-level circulation, the altitudinal gradient, and varying SSTs off the Peruvian coast plays a major role in controlling the precipitation variability in our study area on the western slopes of the central Andes. The major signal in precipitation distribution along the slopes is the positive gradient, which is consistent with Rau et al. (2017) and has a high confidence (R2 = 0.81; Fig. 3a). However, the height dependency of the precipitation weakens during MAM (Fig. 3d; R2 = 0.37), while the relation to the upper-level circulation strengthens as the zonal patterns predominate (Fig. 4b). The explained variance of the leading mode in the EOF analysis is greatest for this season (73%), which indicates the predominance of this pattern. The strong correlation coefficients between the eigenvalues and the rain gauges (0.6–0.9) confirm this coherence that is missing in DJF (0.2–0.5; Table 3). Moreover, the partitioning cluster analysis corroborates these varying atmospheric influences by means of a delineation of local rain regimes along both the altitudinal and meridional gradient (Fig. 5). In particular, in MAM the meridional clustering is pronounced (Fig. 5c) because of the weakening of the BH. This weakening occurs during the transition to the dry season (i.e., MAM), and thus indicates the decreasing impact of the BH with decreasing latitude on precipitation formation (Lenters and Cook 1999; Garreaud 1999). However, in our cluster analysis, we neglected topographical influences such as aspect and exposure and considered solely the temporal distribution of the rain gauges. Nevertheless, their effects are highlighted for DJF and MAM when the insolation is more relevant (Fig. 5b). In contrast, in JJA and SON, only the altitudinal gradient that causes slope-breeze-induced precipitation determines the partitioning, and topographical effects are not featured. Thus, during the transition to the dry season, the position and intensity of the upper-level anticyclone associated with a zonal structure in the circulation are the main drivers in precipitation variability. During the remaining seasons, the impact of the altitudinal gradient is more relevant.

In addition to the local rain regimes we found an inversion of the major precipitation signal (negative gradient) along the three transects T1–T3 (Figs. 1 and 6 and Table 5). These events occur on a daily basis with a duration of up to 6 days. Although Peru is a key region for the ENSO phenomenon (e.g., Aceituno 1988; Vuille et al. 2008; Bourrel et al. 2015), a link to the negative precipitation gradient could not be established. However, a link to the SST values of the Niño-1+2 regions emerged, which is stronger for the negative gradient cases (correlation coefficients between 0.56 and 0.67) than for the positive gradient cases (0.25–0.37). Since the positive gradient is the major precipitation pattern, only a weak relation occurred. Minor differences (T1–T3) in the signal of the negative gradients are caused by the differences in the frequencies of occurrences. Simulations applying the WRF Model (Figs. 710) illustrate the underlying dynamic mechanisms of negative precipitation gradients and reveal that the strength of the locally induced anabatic winds forces the precipitation formation, which was also shown by Garreaud and Aceituno (2001). In POS1 and POS2, the eastern wind direction aloft blocks the western flow regime and causes typical afternoon condensational lifting accompanied by cloud and precipitation formation (Figs. 9a,b). In NEG1, however, the upper-level wind field features a rather weak eastern flow (1–2 m s−1), which blocks the westerlies only partially (Fig. 9c). In the middle troposphere, a strong flow evolves (see Figs. 7 and 9), which affects the boundary layer over the crest and advanced rather cold and dry air masses to the west, as indicated by θe between 334 and 338 K. The easterly flow over the crest converges with thermally induced upslope winds on the western flanks in a conditionally unstable environment because of diabatic heating and thus produces the observed negative gradient. In agreement with previous studies of the central Andes (Garreaud 1999; Garreaud and Aceituno 2001; Vuille et al. 2000a; Lenters and Cook 1999), the development of local upslope winds seems to be strongly affected by the mid- and upper-level flow regime, with easterly wind anomalies modulating the local-scale diurnal circulation by downward mixing of momentum. Further, Lenters and Cook (1999), Garreaud (1999), and Vuille and Keimig (2004) reported that the position and intensity of the BH control the upper-level circulation with a more southerly position favoring enhanced easterlies and an increase in precipitation, and vice versa. In NEG1, we found that the BH has a northerly position relative to that of the positive cases, whereas it is not established in NEG2. The latter causes the similarity between POS2 and NEG2 in terms of the wind speed and direction.

As demonstrated in Fig. 10a, a further component in this interplay is the near-coastal SST, which is also ascertained by the correlation analysis in Table 5. The coastal SSTs are associated with a low-level wind along the Pacific coast favored by upwelling related to the Peru Current as described by Aguirre et al. (2012) and Rahn and Garreaud (2014). Changes in the SST values weaken or strengthen the low-level winds, which then also modify the large-scale subsidence and vertical velocity (Muñoz 2008). The usually low SSTs of the Peruvian coastal areas generate a heat gradient between the ocean and the land surfaces, leading to a thermal land–sea breeze system (e.g., Rutllant et al. 2003; Takahashi 2012). Enfield (1981) showed that the alongshore winds near Lima develop during strong thermal contrasts between sea and land surfaces, which intensifies in conjunction with anomalous SSTs. Strong low-level winds suppress sea breezes and thus, the formation of precipitation along the western slopes. Dewitte et al. (2011) reported a link between the surface stratification, wind stress anomalies, and their effects on SST anomalies along the Peruvian coast. The authors further showed that the low-level winds are related to anomalies in the South Pacific high, as also addressed by Garreaud (1999) and Muñoz (2008). In our study, we also found that colder SSTs in more offshore areas rather than close to the coast (around −1 K; see Fig. 10c) led to strong southeastern low-tropospheric winds embedded in the South Pacific high pressure system, which developed an anticyclonic effect, as described by Lettau (1967) (Fig. 10a). These strong low-level winds farther away off the Peruvian coast in association with downward mixing of momentum result in weaker sea breezes and convection, and thus, less precipitation in lower elevations. In contrast, NEG1 featured a southwestern onshore wind (Fig. 10b) likely because of warmer SSTs over the Pacific (Fig. 10c), which also favor convective activity by the observed moisture-laden atmosphere. Moreover, the sea and land surface contrast near the coast is stronger, which produces stronger land–sea breezes, uplift of air masses and subsequently precipitation. Thus, these variations in the SST patterns off the Peruvian coast and their atmospheric effects account for the positive link to the Niño-1+2 regions SST values and lead to negative precipitation gradients, while the anomalies reveal no signal (Table 5).

As shown, the precipitation variability at the western slopes is a cross-scale interplay of fluctuations in the large-scale atmospheric conditions, breeze-dominated slope flows and varying SST patterns. The dominant signal along the slopes is an increase of precipitation with increasing elevation, but an inversion of this signal on smaller time scales occur. While the large-scale circulation features aloft affects the thermally induced local upslope flows, the strength of southeastern low-level winds related to the SSTs plays a major role through reduced anticyclonic effects accompanied by a blocking of strong easterlies in the middle troposphere. This allows for sufficient moisture availability in the usually dry conditions, which produces the decreasing precipitation patterns with increasing elevation.

The precipitation formation in this area is highly variable and patterns are strongly differentiated in space and time. Moreover, there are complex interacting controls across scale, which highlights the need for further analyses not only on interseasonal but also on intraseasonal time scales.

Acknowledgments

NCEP reanalysis data were provided by the NOAA/OAR/ESRL PSD, Boulder, Colorado, from their website at http://www.esrl.noaa.gov/psd/. The precipitation data were provided by the Peruvian Meteorological and Hydrological National Service (SENAMHI) within the framework of the LiWa-Project (Sustainable Water and Wastewater Management in Urban Growth Centres Coping with Climate Change—Concepts for Lima Metropolitana), which was funded by the German Federal Ministry of Education and Research. The model output of this article is available from the corresponding author upon request (trachtek@staff.uni-marburg.de).

REFERENCES

  • Aceituno, P., 1988: On the functioning of the Southern Oscillation in the South American sector. Part I: Surface climate. Mon. Wea. Rev., 116, 505524, https://doi.org/10.1175/1520-0493(1988)116<0505:OTFOTS>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Aceituno, P., and A. Montecinos, 2000: Diurnal cycle over the South American Altiplano: Comparison of NCEP/NCAR reanalysis with upper-air observations during the Visviri field experiment. Preprints, Sixth Int. Conf. on Southern Hemisphere Meteorology and Oceanography, Santiago, Chile, Amer. Meteor. Soc., 412–413.

  • Aguirre, C., Ó. Pizarro, P. T. Strub, R. Garreaud, and J. A. Barth, 2012: Seasonal dynamics of the near-surface alongshore flow off central Chile. J. Geophys. Res., 117, C01006, https://doi.org/10.1029/2011JC007379.

    • Search Google Scholar
    • Export Citation
  • Barnett, T. P., J. C. Adam, and D. P. Lettenmaier, 2005: Potential impacts of a warming climate on water availability in snow-dominated regions. Nature, 438, 303309, https://doi.org/10.1038/nature04141.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Bourrel, L., and et al. , 2015: Low-frequency modulation and trend of the relationship between ENSO and precipitation along the northern to centre Peruvian Pacific coast. Hydrol. Processes, 29, 12521266, https://doi.org/10.1002/hyp.10247.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Bradley, R. S., M. Vuille, H. F. Diaz, and W. Vergara, 2006: Threats to water supplies in the tropical Andes. Science, 312, 17551756, https://doi.org/10.1126/science.1128087.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Chamorro Chávez, A., 2015: Stochastic and hydrological modelling for climate change prediction in the Lima region, Peru. Ph.D. dissertation, Universität Stuttgart, 119 pp., http://dx.doi.org/10.18419/opus-638.

    • Crossref
    • Export Citation
  • Dewitte, B., S. Illig, L. Renault, K. Goubanova, K. Takahashi, D. Gushchina, K. Mosquera, and S. Purca, 2011: Modes of covariability between sea surface temperature and wind stress intraseasonal anomalies along the coast of Peru from satellite observations (2000–2008). J. Geophys. Res., 116, C04028, https://doi.org/10.1029/2010JC006495.

    • Search Google Scholar
    • Export Citation
  • Drenkhan, F., M. Carey, C. Huggel, J. Seidel, and M. T. Oré, 2015: The changing water cycle: Climatic and socioeconomic drivers of water-related changes in the Andes of Peru. Wiley Interdiscip. Rev.: Water, 2, 715733, https://doi.org/10.1002/wat2.1105.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Eda, L. E. H., and W. Chen, 2010: Integrated water resources management in Peru. Procedia Environ. Sci., 2, 340348, https://doi.org/10.1016/j.proenv.2010.10.039.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Enfield, D. B., 1981: Thermally driven wind variability in the planetary boundary layer above Lima, Peru. J. Geophys. Res., 86, 20052016, https://doi.org/10.1029/JC086iC03p02005.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Garreaud, R. D., 1999: Multiscale analysis of the summertime precipitation over the central Andes. Mon. Wea. Rev., 127, 901921, https://doi.org/10.1175/1520-0493(1999)127<0901:MAOTSP>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Garreaud, R. D., and P. Aceituno, 2001: Interannual rainfall variability over the South American Altiplano. J. Climate, 14, 27792789, https://doi.org/10.1175/1520-0442(2001)014<2779:IRVOTS>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Garreaud, R. D., M. Vuille, and A. C. Clement, 2003: The climate of the Altiplano: Observed current conditions and mechanisms of past changes. Palaeogeogr. Palaeoclimatol. Palaeoecol., 194, 522, https://doi.org/10.1016/S0031-0182(03)00269-4.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Garreaud, R. D., M. Vuille, R. Compagnucci, and J. Marengo, 2009: Present-day South American climate. Palaeogeogr. Palaeoclimatol. Palaeoecol., 281, 180195, https://doi.org/10.1016/j.palaeo.2007.10.032.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Houston, J., and A. J. Hartley, 2003: The central Andean west-slope rainshadow and its potential contribution to the origin of hyper-aridity in the Atacama Desert. Int. J. Climatol., 23, 14531464, https://doi.org/10.1002/joc.938.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Illig, S., and et al. , 2014: Forcing mechanisms of intraseasonal SST variability off central Peru in 2000–2008. J. Geophys. Res. Oceans, 119, 35483573, https://doi.org/10.1002/2013JC009779.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Kain, J. S., 2004: The Kain–Fritsch convective parameterization: An update. J. Appl. Meteor., 43, 170181, https://doi.org/10.1175/1520-0450(2004)043<0170:TKCPAU>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Kalnay, E., and et al. , 1996: The NCEP/NCAR 40-Year Reanalysis Project. Bull. Amer. Meteor. Soc., 77, 437471, https://doi.org/10.1175/1520-0477(1996)077<0437:TNYRP>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Kaufman, L., and P. Rousseeuw, 1987: Clustering by means of medoids. Statistical Data Analysis Based on the L1 -Norm and Related Methods, Y. Dodge, Ed., North-Holland, 405–416.

  • Lavado Casimiro, W. S., J. Ronchail, D. Labat, J. C. Espinoza, and J. L. Guyot, 2012: Basin-scale analysis of rainfall and runoff in Peru (1969–2004): Pacific, Titicaca and Amazonas drainages. Hydrol. Sci. J., 57, 625642, https://doi.org/10.1080/02626667.2012.672985.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Lenters, J. D., and K. H. Cook, 1997: On the origin of the Bolivian high and related circulation features of the South American climate. J. Atmos. Sci., 54, 656678, https://doi.org/10.1175/1520-0469(1997)054<0656:OTOOTB>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Lenters, J. D., and K. H. Cook, 1999: Summertime precipitation variability over South America: Role of the large-scale circulation. Mon. Wea. Rev., 127, 409431, https://doi.org/10.1175/1520-0493(1999)127<0409:SPVOSA>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Lettau, H. H., 1967: Small to large-scale features of boundary layer structure over mountain slopes. Proc. Symp. on Mountain Meteorology, Vol. 26, Fort Collins, CO, Department of Atmospheric Science, Colorado State University, 1–73.

  • Lynch, B. D., 2012: Vulnerabilities, competition and rights in a context of climate change toward equitable water governance in Peru’s Rio Santa Valley. Global Environ. Change, 22, 364373, https://doi.org/10.1016/j.gloenvcha.2012.02.002.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Ma, H.-Y., C. R. Mechoso, H. Xiao, C.-M. Wu, Y. Xue, and F. De Sales, 2010: Connection between the South Pacific anti-cyclone, Peruvian stratocumulus, and the South American monsoon system. CLIVAR Exchanges, Vol. 15, International CLIVAR Project Office, Southampton, United Kingdom, 569–572.

  • McGlone, D., and M. Vuille, 2012: The associations between El Niño–Southern Oscillation and tropical South American climate in a regional climate model. J. Geophys. Res., 117, D06105, https://doi.org/10.1029/2011JD017066.

    • Search Google Scholar
    • Export Citation
  • Mlawer, E. J., S. J. Taubman, P. D. Brown, M. J. Iacono, and S. A. Clough, 1997: Radiative transfer for inhomogeneous atmospheres: RRTM, a validated correlated-k model for the longwave. J. Geophys. Res., 102, 16 66316 682, https://doi.org/10.1029/97JD00237.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Mohr, K. I., D. Slayback, and K. Yager, 2014: Characteristics of precipitation features and annual rainfall during the TRMM era in the central Andes. J. Climate, 27, 39824001, https://doi.org/10.1175/JCLI-D-13-00592.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Muñoz, R., 2008: Diurnal cycle of surface winds over the subtropical southeast Pacific. J. Geophys. Res., 113, D13107, https://doi.org/10.1029/2008JD009957.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Nakanishi, M., and H. Niino, 2006: An improved Mellor–Yamada level-3 model: Its numerical stability and application to a regional prediction of advection fog. Bound.-Layer Meteor., 119, 397407, https://doi.org/10.1007/s10546-005-9030-8.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Neukom, R., M. Rohrer, P. Calanca, N. Salzmann, C. Huggel, D. Acuña, D. A. Christie, and M. S. Morales, 2015: Facing unprecedented drying of the central Andes? Precipitation variability over the period AD 1000–2100. Environ. Res. Lett., 10, 084017, https://doi.org/10.1088/1748-9326/10/8/084017.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • OCHA, 2017: Peru: rainy season as of 10 May 2017. U.N. Office for the Coordination of Humanitarian Affairs Situation Rep. 10, 15 pp., https://reliefweb.int/sites/reliefweb.int/files/resources/redhum-pe-informe_de_situacion_10-eng-ocha-20170512.pdf.

  • Perry, L. B., A. Seimon, and G. M. Kelly, 2014: Precipitation delivery in the tropical high Andes of southern Peru: New findings and paleoclimatic implications. Int. J. Climatol., 34, 197215, https://doi.org/10.1002/joc.3679.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Perry, L. B., and et al. , 2017: Characteristics of precipitating storms in glacierized tropical Andean cordilleras of Peru and Bolivia. Ann. Amer. Assoc. Geogr., 107, 309322, https://doi.org/10.1080/24694452.2016.1260439.

    • Search Google Scholar
    • Export Citation
  • Rahn, D. A., and R. D. Garreaud, 2014: A synoptic climatology of the near-surface wind along the west coast of South America. Int. J. Climatol., 34, 780792, https://doi.org/10.1002/joc.3724.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Rasmussen, K. L., and R. A. Houze Jr., 2011: Orogenic convection in subtropical South America as seen by the TRMM satellite. Mon. Wea. Rev., 139, 23992420, https://doi.org/10.1175/MWR-D-10-05006.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Rau, P., L. Bourrel, D. Labat, P. Melo, B. Dewitte, F. Frappart, W. Lavado, and O. Felipe, 2017: Regionalization of rainfall over the Peruvian Pacific slope and coast. Int. J. Climatol., 37, 143158, https://doi.org/10.1002/joc.4693.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Rayner, N. A., D. E. Parker, E. B. Horton, C. K. Folland, L. V. Alexander, D. P. Rowell, E. C. Kent, and A. Kaplan, 2003: Global analysis of sea surface temperature, sea ice, and night marine air temperature since the late nineteenth century. J. Geophys. Res., 108, 4407, https://doi.org/10.1029/2002JD002670.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Rodwell, M. J., and B. J. Hoskins, 2001: Subtropical anticyclones and summer monsoons. J. Climate, 14, 31923211, https://doi.org/10.1175/1520-0442(2001)014<3192:SAASM>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Rutllant, J. A., H. Fuenzalida, and P. Aceituno, 2003: Climate dynamics along the arid northern coast of Chile: The 1997–1998 Dinámica del Clima de la Región de Antofagasta (DICLIMA) experiment. J. Geophys. Res., 108, 4538, https://doi.org/10.1029/2002JD003357.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Salzmann, N., C. Huggel, S. U. Nussbaumer, and G. Ziervogel, 2016: Setting the scene: Adapting to climate change—A large-scale challenge with local-scale impacts. Climate Change Adaptation Strategies—An Upstream–Downstream Perspective, N. Salzmann et al., Eds., Springer, 3–16.

    • Crossref
    • Export Citation
  • Schütze, M., 2015: Sustainable water and wastewater management in urban growth centres coping with climate change—Concepts for Lima Metropolitana (Perú). BMBF Final Rep., 55 pp., http://www.lima-water.de/en/documents.html?Menu=5.

  • Silva Dias, P. L., W. H. Schubert, and M. DeMaria, 1983: Large-scale response of the tropical atmosphere to transient convection. J. Atmos. Sci., 40, 26892707, https://doi.org/10.1175/1520-0469(1983)040<2689:LSROTT>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Skamarock, W. C., and J. B. Klemp, 2008: A time-split nonhydrostatic atmospheric model for weather research and forecasting applications. J. Comput. Phys., 227, 34653485, https://doi.org/10.1016/j.jcp.2007.01.037.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Takahashi, K., 2005: The annual cycle of heat content in the Peru Current region. J. Climate, 18, 49374954, https://doi.org/10.1175/JCLI3572.1.

  • Takahashi, K., 2012: Thermotidal and land-heating forcing of the diurnal cycle of oceanic surface winds in the eastern tropical Pacific. Geophys. Res. Lett., 39, L04805, https://doi.org/10.1029/2011GL050692.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Tewari, M., and et al. , 2004: Implementation and verification of the unified NOAH land surface model in the WRF model. 20th Conf. on Weather Analysis and Forecasting/16th Conf. on Numerical Weather Prediction, Seattle, WA, Amer. Meteor. Soc., 14.2A, https://ams.confex.com/ams/pdfpapers/69061.pdf.

  • Thompson, G., R. M. Rasmussen, and K. Manning, 2004: Explicit forecasts of winter precipitation using an improved bulk microphysics scheme. Part I: Description and sensitivity analysis. Mon. Wea. Rev., 132, 519542, https://doi.org/10.1175/1520-0493(2004)132<0519:EFOWPU>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Vera, C., and et al. , 2006: Toward a unified view of the American monsoon systems. J. Climate, 19, 49775000, https://doi.org/10.1175/JCLI3896.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Viviroli, D., H. H. Dürr, B. Messerli, M. Meybeck, and R. Weingartner, 2007: Mountains of the world, water towers for humanity: Typology, mapping, and global significance. Water Resour. Res., 43, W07447, https://doi.org/10.1029/2006WR005653.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Vuille, M., and F. Keimig, 2004: Interannual variability of summertime convective cloudiness and precipitation in the central Andes derived from ISCCP-B3 data. J. Climate, 17, 33343348, https://doi.org/10.1175/1520-0442(2004)017<3334:IVOSCC>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Vuille, M., R. S. Bradley, and F. Keimig, 2000a: Climate variability in the Andes of Ecuador and its relation to tropical Pacific and Atlantic sea surface temperature anomalies. J. Climate, 13, 25202535, https://doi.org/10.1175/1520-0442(2000)013<2520:CVITAO>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Vuille, M., R. S. Bradley, and F. Keimig, 2000b: Interannual climate variability in the central Andes and its relation to tropical Pacific and Atlantic forcing. J. Geophys. Res., 105, 12 44712 460, https://doi.org/10.1029/2000JD900134.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Vuille, M., G. Kaser, and I. Juen, 2008: Glacier mass balance variability in the Cordillera Blanca, Peru and its relationship with climate and the large-scale circulation. Global Planet. Change, 62, 1428, https://doi.org/10.1016/j.gloplacha.2007.11.003.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Wang, C., and P. C. Fiedler, 2006: ENSO variability and the eastern tropical Pacific: A review. Prog. Oceanogr., 69, 239266, https://doi.org/10.1016/j.pocean.2006.03.004.

    • Crossref
    • Search Google Scholar
    • Export Citation
Save