Full Terms & Conditions of access and use can be found at
http://www.tandfonline.com/action/journalInformation?journalCode=thsj20
Hydrological Sciences Journal
ISSN: 0262-6667 (Print) 2150-3435 (Online) Journal homepage: http://www.tandfonline.com/loi/thsj20
Basin-scale analysis of rainfall and runoff in
Peru (1969–2004): Pacific, Titicaca and Amazonas
drainages
Waldo Sven Lavado Casimiro , Josyane Ronchail , David Labat , Jhan Carlo
Espinoza & Jean Loup Guyot
To cite this article: Waldo Sven Lavado Casimiro , Josyane Ronchail , David Labat , Jhan
Carlo Espinoza & Jean Loup Guyot (2012) Basin-scale analysis of rainfall and runoff in Peru
(1969–2004): Pacific, Titicaca and Amazonas drainages, Hydrological Sciences Journal, 57:4,
625-642, DOI: 10.1080/02626667.2012.672985
To link to this article:  https://doi.org/10.1080/02626667.2012.672985
Published online: 03 Apr 2012.
Submit your article to this journal 
Article views: 1170
Citing articles: 30 View citing articles 
625Hydrological Sciences Journal – Journal des Sciences Hydrologiques, 57(4) 2012
Basin-scale analysis of rainfall and runoff in Peru (1969–2004): Pacific,
Titicaca and Amazonas drainages
Waldo Sven Lavado Casimiro1,2,3 , Josyane Ronchail4, David Labat2, Jhan Carlo Espinoza5
and Jean Loup Guyot2,4
1Servicio Nacional de Meteorología e Hidrología, SENAMHI, Casilla 11 1308, Lima 11, Peru
wlavado@senamhi.gob.pe
2Geosciences Environnement Toulouse (CNRS, IRD, Université de Toulouse, OMP), 14 Avenida Edouard Belin, F-31400 Toulouse,
France
3Universidad Agraria La Molina (UNALM). Av. Universidad s/n La Molina, Lima 12, Perú
4Université Paris Denis Diderot, Sorbonne Paris Cité, F-75013 Paris, France
5Instituto Geofísico del Perú (IGP), Calle Badajoz 169, Mayorazgo IV Etapa, Lima 03, Peru
Received 12 July 2010; accepted 25 August 2011; open for discussion until 1 November 2012
Editor Z.W. Kundzewicz; Associate editor Š. Blažková
Citation Lavado C., W.S., Ronchail, J., Labat, D., Espinoza, J.C. and Guyot, J.L., 2012. Basin-scale analysis of rainfall and runoff in
Peru (1969–2004): Pacific, Titicaca and Amazonas watersheds. Hydrological Sciences Journal, 57 (4), 625–642.
Abstract According to the Peruvian agricultural ministry, the Pacific watersheds where the great cities and intense
farming are located only benefit from 1% of the available freshwater in Peru. Hence a thorough knowledge of the
hydrology of this region is of particular importance. In the paper, analysis of this region and of the two other main
Peruvian drainages, the Titicaca and Amazonas are reported. Rainfall and runoff data collected by the Peruvian
National Service of Meteorology and Hydrology (SENAMHI) and controlled under the Hydrogeodynamics of the
Amazon Basin (HyBAm) project is the basis of this basin-scale study that covers the 1969–2004 period. Beyond
the strong contrasting rainfall conditions that differentiate the dry coastal basins and the wet eastern lowlands,
details are given about in situ runoff and per basin rainfall distribution in these regions, and about their different
altitude–rainfall relationships. Rainfall and runoff variability is strong in the coastal basins at seasonal and inter-
annual time scales, and related to extreme El Niño events in the Pacific Ocean. However, rainfall and runoff are
more regular in the Andes and Amazonas at the inter-annual time scale. Warm sea-surface temperatures in the
northern tropical Atlantic tend to produce drought in the southern Andes basins. Moreover, significant trends and
change-points are observed in the runoff data of Amazonas basins where rainfall and runoff decrease, especially
after the mid-1980s and during the low-stage season. Almost all the coastal basins show some change in minimum
runoff during the last 35 years while no change is observed in rainfall. This means that human activity may have
changed runoff in this region of Peru, but this hypothesis deserves more study.
Key words rainfall; runoff; Titicaca basin; Amazon; Pacific coast; Peru; tropical Atlantic
Analyse de la pluie et de l’écoulement au Pérou (1969–2004) : bassins versants du Pacifique, du
Lac Titicaca et de l’Amazone
Résumé D’après le Ministère péruvien de l’agriculture, les bassins du versant Pacifique, qui concentrent les
grandes villes et où se pratique une agriculture intensive, ne bénéficient que de 1% de l’eau douce disponible
au Pérou. C’est pourquoi cet article présente l’hydrologie de cette région, dont la connaissance approfondie
est d’une importance capitale pour l’économie du Pérou. Les deux autres grands bassins versants péruviens,
celui, endoréique, du Titicaca et celui de l’Amazone, seront également étudiés. Les données mensuelles de
précipitations et de débits pour la période de 1969–2004 à l’échelle du bassin versant, sont mises à dispo-
sition par le Service national de météorologie et d’hydrologie (SENAMHI) de Pérou, et elles sont analysées
dans le cadre du programme « Hydrogéodynamique du bassin amazonien » (HyBAm). Au-delà du con-
traste classique entre bassins versants côtiers secs et plaines orientales humides, la répartition spatiale de
l’écoulement et des pluies est détaillée avec, notamment, une régionalisation des liens entre pluie et altitude. Une
variabilité temporelle élevée des précipitations et de l’écoulement est mise en évidence dans les bassins côtiers,
ISSN 0262-6667 print/ISSN 2150-3435 online
© 2012 IAHS Press
http://dx.doi.org/10.1080/02626667.2012.672985
http://www.tandfonline.com
626 Waldo Sven Lavado Casimiro et al.
à des échelles de temps saisonnière et interannuelle, en relation avec les événements extrêmes El Niño dans l’Océan
Pacifique. Par contre, les précipitations et l’écoulement apparaissent comme moins variables dans les Andes ori-
entales et en Amazonie au pas de temps interannuel. Cependant, des températures de surface de l’océan élevées
dans le nord de l’Atlantique tropical ont tendance à produire des sécheresses dans les bassins andins méridionaux.
Par ailleurs, des tendances et des ruptures significatives dans la moyenne sont observées en Amazonie où pluie et
écoulement diminuent, surtout depuis le milieu des années 1980, durant la saison d’étiage. Dans presque tous les
bassins côtiers, l’écoulement d’étiage est sujet à des tendances diverses (à la hausse ou à la baisse), au cours des
35 dernières années bien que de telles tendances ne soient pas visibles sur les séries de précipitations. Cela signifie
que l’activité humaine peut être responsable de cette évolution sur la côte péruvienne. La confirmation de cette
hypothèse nécessite des études complémentaires.
Mots clefs pluie; écoulement; bassin du Lac Titicaca; Amazonie; côte Pacifique; Pérou; Atlantique tropical
INTRODUCTION
The north–south Andes cordillera divides Peru
(1 285 216 km2) into three main drainages (Fig. 1),
one draining towards the Pacific Ocean (Pacific
drainage—Pd), another towards the Amazon basin
(Amazonas drainage—Ad) and the southern,
endorheic Lake Titicaca basin on the Altiplano
(Titicaca drainage—Td). The Peruvian National
Water Agency reports that the Pacific, Titicaca and
Amazon drainages represent, respectively, 21.7%,
3.8% and 74.5% of the Peruvian territory and
include 62, 13 and 84 main sub-basins, respectively
(Ruiz et al. 2008). The multi-annual water balance
(difference between rainfall and evapotranspiration)
computed for the period 1969–1999 by UNESCO
(2006) indicates that the available surface water
is 16.4 mm year-1 in Pd, 129.8 mm in Td and
2696.6 mm in Ad. Unfortunately, the accessibility
of freshwater resources in Peru is the converse of
population density; 88% of the population lives
along the Pacific coast, around Lake Titicaca and
in the Andean zones of the Amazon basin, where
only 2% of the Peruvian freshwater resources are
available. Moreover, the greatest Peruvian cities
(i.e. Lima, Arequipa and Piura) are in Pd with only
0.6% of the available freshwater in Peru. According
to the agricultural ministry (http://www.minag.gob.
pe), agriculture along the Pacific coast has been
developed thanks to irrigation that uses about 80% of
the available water, while domestic usage consumes
about 12%. Private and public investments have been
used in order to develop this infrastructure aimed at
increasing agricultural exports. Water resources are
also essential in the other drainages. For instance, a
large rural population lives around Lake Titicaca and
depends on water for farming produce. Moreover,
transport activities in the Amazon basin are extremely
dependent on rivers, due to the absence of roads in the
dense forest, and transport services are interrupted as
a result of extreme low water levels.
Most studies of Peruvian hydrology have focused
on only one of the main drainages and seldom
on all three of them. The aim of this study is
to propose a synoptic outline of the mean charac-
teristics of rainfall and discharge in Peru and of
their recent variability, using a comprehensive set
of rainfall and runoff data that covers the whole
Peruvian territory (Pacific, Amazonas and Titicaca
drainages). This is made possible thanks to the
collection and the validation of information by
the Peruvian National Meteorology and Hydrology
Service (Servicio Nacional de Meteorología e
Hidrología, SENAMHI, http://www.senamhi.gob.pe)
and, in particular, the collection of discharge infor-
mation through runoff gauging in big rivers of the
Peruvian Amazon under the Hydrogeodynamics of
the Amazon Basin (HyBAm, http://www.ore-hybam.
org/) project, a collaboration between SENAMHI and
the French Institute of Research for Development
(IRD, http://www.ird.fr/).
THE PERUVIAN DRAINAGES
Peruvian hydroclimatology is influenced by the dis-
ruption of the large-scale circulation patterns caused
by the Andes cordillera, the contrasting oceanic
boundary conditions and the landmass distribution
(Garreaud et al. 2009).
The Pd features small basins with rivers flow-
ing east to west, from the Andes toward the Pacific
Ocean (Fig. 1). These basins have bare and steep
slopes that favour significant erosion and flooding
during very rainy episodes. Weak rainfall along the
Pacific coast is related to the large-scale mid tropo-
spheric subsidence over the southeastern subtropical
Pacific Ocean, enhanced by the coastal upwelling of
cold water. The warming of the air above and the cold
sea-surface temperature (SST) result in a cool, moist
marine boundary layer of 500–1000 m thickness,
capped by a strong temperature inversion (Garreaud
Basin-scale analysis of rainfall and runoff in Peru (1969–2004): Pacific, Titicaca and Amazonas drainages 627
(a)
(c)
(b)
Fig. 1 (a) Elevation of the study area, main rivers, limits of the selected basins in the Pacific, Titicaca and Amazon drainages
and location of the runoff gauges (see Table 1). (b) Location of the study zone in South America. (c) Location of the rainfall
gauges used in this study.
et al. 2002). In the austral summer, the slight weaken-
ing of the southeast Pacific anticyclone and the south-
ward displacement of the Pacific intertropical conver-
gence zone (ITCZ) allow the development of a rainy
season. UNESCO (2006) assesses that mean annual
rainfall and runoff in Pd are very low, about 274 and
168 mm, respectively, during the 1969–1999 period.
However, within the region, rainfall and runoff show
significant internal differences. Indeed, rainfall is
more abundant along the northern coast and declines
towards the south where conditions of extreme arid-
ity occur due to the large-scale subsidence acting in
concert with regional factors (Rutllant et al. 2003,
Garreaud et al. 2009).
The Ad is characterized by hot tropical lowlands
at near sea level and freezing cold, tropical, high
mountains peaking at more than 6000 m a.s.l. ele-
vation. The rivers in the Ad have steep slopes in the
western Andean region and very small gradients in
the flat eastern lowlands; seldom do such dramatic
environmental contrasts lie closer together. Naturally,
each of these two realms is subject to different, but in
many aspects interrelated, sets of physical, geochem-
ical and biological parameters (ACTO 2005). Very
wet conditions are observed in the Amazon basin,
due to the convection of moist and hot air that orig-
inates from the evapotranspiration of the rainforest
and the advection of humid air from the tropical
Atlantic. Rainfall is more abundant during the South
American monsoon (Zhou and Lau 1998), i.e. in the
austral summer, in the southernmost tropical regions
and around the autumn equinox towards the Equator,
where the seasonal variability is nevertheless very low
(Johnson 1976, Espinoza et al. 2009b). According
to UNESCO (2006), mean rainfall over Ad during
the 1969–1999 period was close to 2060 mm year-1,
but important spatial variations are observed with
a general rainfall decrease toward the tropics (i.e.
the south) and toward the Andes as rainfall tends to
diminish with altitude, to <1500 mm above 2000 m
a.s.l. (Espinoza et al. 2009b). However, in the Andes,
very high and low rainfall values (between 6000 and
250 mm/year) can be observed at nearby stations due
to the leeward or windward position of raingauges
(Espinoza et al. 2009b).
The Td is located on the Altiplano (south-
east of Peru, Fig. 1) and its endorheic drainage
is realized from high Andean summits (surpassing
6000 m a.s.l.) with streams flowing towards Lake
Titicaca (∼3800 m a.s.l.), then the Desaguadero
River and finally Poopo and Salar de Coipasa lakes
in Bolivia. UNESCO (2006) assesses that, on the
Altiplano, mean annual rainfall and runoff during
the 1969–1999 period were 813 and 89 mm, respec-
tively. The rainy season is in the austral summer,
with a peak in January, and is characterized by
628 Waldo Sven Lavado Casimiro et al.
intense convective activity combined with moisture
advection from the Amazon basin (Aceituno and
Montecinos 1993, Chaffaut et al. 1998, Vuille et al.
1998, Garreaud 1999, Garreaud et al. 2003, Vizy
and Cook 2007). During the austral winter, the moist
eastern flux is replaced by westerly winds, which pro-
vide dry air related to the atmospheric stability over
the Pacific Ocean. This strong seasonality explains
the moderate amounts of annual rainfall. However,
locally, around Lake Titicaca, water vapour is more
abundant and annual rainfall exceeds 1000 mm/year
(Roche et al. 1990).
The El Niño Southern Oscillation (ENSO) has
opposing influences on inter-annual variability of
rainfall and runoff in the northern Pacific drainage
and in the Lake Titicaca region (Waylen and Caviedes
1986, Ropelewski and Halpert 1987, Aceituno 1988,
Tapley and Waylen 1989, 1990, Rome-Gaspaldy and
Ronchail 1998, Vuille et al. 2000, Waylen and Poveda
2002, Romero et al. 2007, among others). During El
Niño, rainfall and runoff are higher than normal in the
northern Pd, while it is associated with hydrological
drought in Td and over the high Andes. Dramatic
rainfall on the coast during El Niño is related to the
inversion of the Pacific Walker cell, with enhanced
ascendance over the unusually warm waters in the
central and eastern Pacific. On the Altiplano, the El
Niño drought is associated with a warming of the
tropical atmosphere, a strengthening of the wester-
lies and less advection of moist air from the Amazon
basin (Aceituno and Montecinos 1993, Garreaud and
Aceituno 2001). Moreover, inter-annual rainfall vari-
ability in the southern Amazon basin is negatively
related to North Atlantic sea-surface temperatures
(NATL SSTs) (Ronchail et al. 2002, Yoon and Zeng
2009, Lavado 2010), as well as runoff in the Peruvian
Amazonas basin (Espinoza et al. 2009a). Warm SST
conditions over the northern tropical Atlantic favour
convection in this oceanic region and the converse,
subsidence, over the southern tropics.
Long-term rainfall and runoff variability have
also been observed in Peru. Marengo (1995, 1998)
found decreasing runoff trends during the 20th cen-
tury at some gauging stations located in the northern
Pd. These were not natural, being often related to
water abstraction for irrigation and domestic usage
in the growing cities and farming zones. In contrast,
runoff of the high-elevation, upstream Santa River has
becomemore regular as its basin is undergoing glacier
retreat (Kaser et al. 2003, Mark and Seltzer 2003,
Mark et al. 2005, Pouyaud et al. 2005, Vuille et al.
2008a, 2008b, among others). On the Altiplano, a
13-year climatic periodicity is observed in Quelccaya
ice-cores (Melice and Roucou 1998) and a recent
decrease (since the end of the 1980s) is detected
in the level of Lake Titicaca, which acts as a rain-
fall gauge for the Td region (Rigsby et al. 2003).
In the Ad, Gentry and Lopez-Parodi (1980) associated
upward (downward) trends in maximum (minimum)
water levels in Iquitos (AQ-2, see Table 1) during the
1962–1978 period with deforestation, as no clear rain-
fall change had been observed during the same period.
Nordin et al. (1982) and Richey et al. (1989) dis-
cussed these results, as did Rocha et al. (1989), who
found positive rainfall trends in Iquitos and Pucallpa
stations during the 1955–1979 and 1957–1981 peri-
ods, respectively. Marengo (2004) and Espinoza et al.
(2006, 2007, 2009a) suggest that the long-term rain-
fall evolutions in the north and the south of the basin
are in opposition with the Pacific Decadal Oscillation.
During the 1990–2005 period, Espinoza et al. (2009a)
and Espinoza et al. (2007), respectively, describe
an upward runoff trend in the northern Ad (AQ-3,
see Table 2), and a downward runoff trend over the
southern Ad (AQ-5, see Table 2).
DATA AND METHODS
This work is based on monthly rainfall and dis-
charge data in 33 basins (29 in Pd, three in Td and
two in Ad), during the 1969/70–2003/04 period,
with the hydrological year running from September to
August (Table 1, Fig. 1). Basin subdivision was made
using the Digital ElevationModel (DEM) provided by
the National Aeronautics and Space Administration
(NASA) through the Shuttle Radar Topography
Mission (SRTM – www2.jpl.nasa.gov/srtm). The
SRTM data are available at 3 arc second (∼90-m res-
olution). A description of this model is provided by
Farr et al. (2007).
Monthly rainfall data for 1965 to 2007 are
from the SENAMHI network and have been checked
by this institute. The Pd and Td data have
also been checked by UNESCO (2006), for the
1969–1999 period. Under the HyBAm project, rain-
fall data over Ad underwent preliminary qual-
ity control by Espinoza et al. (2009b) for the
1965–2003 period, and by Lavado (2010) for the
2003–2007 period, using the Regional Vector Method
(RVM; Hiez 1977, Brunet-Moret 1979, and see the
detailed presentation of the method in Espinoza et al.
2009b). Rainfall stations are more abundant in Pd and
Td (119 stations) and, taking into account its large
surface area, less numerous in the Ad (48 stations).
Basin-scale analysis of rainfall and runoff in Peru (1969–2004): Pacific, Titicaca and Amazonas drainages 629
Table 1 Gauging and rainfall stations used in this study. Lat.: latitude; Long.: longitude; Alt.: altitude (m a.s.l.); Dr. area:
drainage area (km2) and Qsp.: specific runoff (L s-1 km-2).
Gauging station River Code Lat. Long. Alt. (m a.s.l.) Dr. area (km2) Qsp. (L s-1 km-2)
El Tigre Tumbes PQ-1 3.72◦S 80.47◦W 40 4 802 23.7
El Ciruelo Chira PQ-2 4.30◦S 80.15◦W 250 7 760 14.8
Pte. Ñacara Piura PQ-3 5.11◦S 80.17◦W 119 4 765 5.9
Racarumi Chancay-
Lambayeque
PQ-4 6.63◦S 79.32◦W 250 2 401 14.2
Batan Zaña PQ-5 6.80◦S 79.29◦W 260 681 11.7
Yonan Jequetepeque PQ-6 7.25◦S 79.10◦W 428 3 354 8.3
Salinar Chicama PQ-7 7.67◦S 78.97◦W 350 3 651 6.8
Quirihuac Moche PQ-8 8.08◦S 78.87◦W 200 1 918 4.7
Huacapongo Viru PQ-9 8.38◦S 78.67◦W 280 941 4.3
Pte. Carretera Santa PQ-10 8.97◦S 78.63◦W 18 11 869 16.9
Yanapampa Pativilca PQ-11 10.67◦S 77.58◦W 800 4 270 10.1
Sayan Huaura PQ-12 11.12◦S 77.18◦W 650 2 896 10.0
Santo Domingo Chancay-Huaral PQ-13 11.38◦S 77.05◦W 697 1 881 9.6
Larancocha Chillon PQ-14 11.68◦S 76.80◦W 120 1 238 4.8
Chosica Rimac PQ-15 11.93◦S 76.69◦W 906 2 339 13.3
La Capilla Mala PQ-16 12.52◦S 76.50◦W 424 2 141 7.0
Socsi Cañete PQ-17 13.03◦S 76.20◦W 330 6 003 8.2
Conta San juan PQ-18 13.45◦S 75.98◦W 350 3 144 3.2
Letrayoc Pisco PQ-19 13.65◦S 75.72◦W 720 3 107 6.8
Los Molinos Ica PQ-20 13.92◦S 75.67◦W 460 2 154 0.5
Bella Union Acari PQ-21 15.48◦S 74.63◦W 70 4 369 3.2
Puente Jaqui Yauca PQ-22 15.48◦S 74.45◦W 214 4 245 2.1
Pte. Ocoña Ocoña PQ-23 16.42◦S 73.12◦W 122 16 646 4.3
Huatiapa Majes PQ-24 16.00◦S 72.47◦W 699 13 651 6.3
Pte. Del Diablo Chili PQ-25 16.41◦S 71.50◦W 236 8 750 1.5
La Pascana Tambo PQ-26 16.99◦S 71.64◦W 281 12 884 2.2
Pte. Viejo Locumba PQ-27 17.62◦S 70.77◦W 550 3 639 0.8
La Tranca Sama PQ-28 17.73◦S 70.48◦W 620 1 993 1.5
Aguas Calientes Caplina PQ-29 17.85◦S 70.12◦W 130 569 1.8
Pte. Ramis Ramis TQ-1 15.26◦S 69.87◦W 385 16 229 4.7
Pte. Huancane Huancane TQ-2 15.22◦S 69.79◦W 386 3 714 5.4
Pte. Ilave Ilave TQ-3 16.09◦S 69.63◦W 385 8 714 4.5
Tabatinga Amazonas AQ-1 4.25◦S 69.93◦W 60 890 308 42.7
Tamshiyacu Amazonas AQ-2 4.00◦S 73.16◦W 105 733 596 44.4
San Regis Marañon AQ-3 4.51◦S 73.95◦W 80 359 910
Borja Marañon AQ-4 4.47◦S 77.55◦W 450 115 478
Requena Ucayali AQ-5 5.03◦S 73.83◦W 200 354 316
In the Amazon basin, they are concentrated in the
Andes head basins that are more easily accessible
than the lowlands. The kriging interpolation method,
which establishes a variogram for each spatial point,
was applied to compute the average rainfall in each
basin (Oliver and Webster 1990, Deutsch and Journel
1992). The variograms evaluate the influence of the
16 closest stations according to their distance. This
method was selected because it takes into consider-
ation a possible spatial data gradient which is very
important in a mountainous region.
Monthly runoff data (1969–2004) collected by
the SENAMHI in Peru, and by the National Water
Agency of Brazil (ANA, http://www.ana.gov.br) for
the Tabatinga station (AQ-1), were used. Although
it is the largest part of Peru, very few stations are
located in Ad and long time series are only available
for very large basins that reach almost 1 million km2.
More detailed information about smaller basins is not
useful in the present context as their measurements
only began during the 1980s or 1990s. In contrast, the
basins in Pd and Td are smaller, <20 000 km2, and
have been densely equipped for a long time.
Runoff data homogenization was performed
by SENAMHI using double-cumulative techniques.
Cross-correlations between all stations were calcu-
lated in order to detect highly-correlated stations (at
the 99% level) and to fill missing data using lin-
ear regression, when less than 5% of the data of a
station was missing. On the coast (Pd), two stations
on the Santa (PQ-10) and the Ocoña (PQ-23) are
downstream barrages, but runoff in all the rivers is
630 Waldo Sven Lavado Casimiro et al.
Table 2 Average values and variability of per basin rainfall and runoff, using the hydrological year (September–August),
during the 1969–2004 period (36 years), except for PQ-10 (1969–1999, 31-year period for runoff). P: rainfall; Q: runoff;
MAVC: inter-annual variation coefficients; SVC: seasonal variation coefficients and r: linear correlation coefficient between
rainfall and runoff. Significant values (95% confidence level) are in bold. ∗ rainfall only has been analysed. ∗∗ not computed.
Code P
(mm)
Q
(m3 s-1)
Qmax
(m3 s-1)
Qmin
(m3 s-1)
MAVC
(P)
MAVC
(Q)
MAVC
(Qmax)
MAVC
(Qmin)
r SVC
P Q
PQ-1 917 114 371 15 0.4 0.6 0.5 0.3 0.77 0.8 1.0
PQ-2 1011 115 321 24 0.4 0.6 0.7 0.6 0.57 0.7 0.8
PQ-3 618 28 120 0 0.8 1.5 1.2 3.3 0.96 1.3 1.3
PQ-4 722 34 91 7 0.4 0.3 0.4 0.4 0.79 0.7 0.7
PQ-5 750 8 22 2 0.7 0.6 0.9 0.4 0.89 0.8 0.7
PQ-6 681 28 97 2 0.9 0.6 0.6 0.7 0.75 0.9 1.0
PQ-7 744 25 113 2 0.4 1.0 1.1 0.8 0.93 0.9 1.2
PQ-8 517 9 35 0 0.4 0.9 1.1 1.2 0.90 0.9 1.1
PQ-9 583 4 19 0 0.4 1.3 1.3 1.2 0.68 0.7 1.4
PQ-10 489 201 550 49 0.2 0.4 0.4 0.3 0.36 0.7 0.7
PQ-11 521 43 115 12 0.3 0.3 0.4 0.5 0.59 0.9 0.6
PQ-12 493 29 85 10 0.3 0.4 0.6 0.3 0.61 0.9 0.8
PQ-13 460 18 61 5 0.3 0.4 0.6 0.4 0.75 1.1 0.9
PQ-14 455 6 18 2 0.3 0.4 0.4 0.9 0.47 1.1 0.8
PQ-15 515 31 70 17 0.2 0.3 0.4 0.3 0.52 1.0 0.5
PQ-16 409 15 59 1 0.3 0.5 0.5 0.6 0.65 1.1 1.2
PQ-17 407 49 155 9 0.2 0.5 0.7 0.3 0.64 1.0 0.9
PQ-18 184 10 45 0 0.3 0.7 0.9 1.2 0.57 1.3 1.2
PQ-19 500 21 82 2 0.3 0.5 0.7 0.7 0.44 0.9 1.1
PQ-20 357 1 3 0 0.4 0.5 0.5 ∗∗ 0.43 1.1 1.4
PQ-21 240 14 72 0 0.4 0.8 0.8 3.0 0.49 1.1 1.5
PQ-22 237 9 45 0 0.4 0.8 0.9 0.4 0.50 1.3 1.4
PQ-23 487 71 244 17 0.3 0.5 0.5 0.6 0.50 1.1 0.9
PQ-24 463 86 291 26 0.3 0.3 0.4 0.2 0.75 1.1 0.9
PQ-25 305 13 45 6 0.3 0.7 1.0 0.6 0.64 1.2 0.7
PQ-26 281 28 90 11 0.4 0.6 1.0 0.5 0.62 1.3 0.7
PQ-27 230 3 5 2 0.4 0.2 0.5 0.3 0.40 1.4 0.3
PQ-28 164 3 12 1 0.5 0.6 0.7 1.0 0.82 1.5 1.2
PQ-29 144 1 2 0 0.5 0.3 0.6 0.3 0.65 1.4 0.5
TQ-1 773 76 260 8 0.2 0.3 0.3 0.4 0.82 0.9 1.1
TQ-2 729 20 78 2 0.2 0.4 0.5 0.5 0.86 0.8 1.1
TQ-3 620 39 172 5 0.2 0.5 0.6 0.5 0.78 0.9 1.2
AQ-1 1913 38044 53710 22545 0.1 0.1 0.1 0.2 0.82 0.2 0.3
AQ-2 1701 32605 49238 16603 0.1 0.1 0.1 0.2 0.60 0.2 0.3
AQ-3∗ 1747 0.1 0.5
AQ-4∗ 1463 0.1 0.3
AQ-5∗ 1496 0.1 0.2
perturbed by small water captures. In the Rímac val-
ley, discharge has been corrected, taking off the flow
captured from the Mantaro River in the Amazon
basin. The “natural” Rímac discharge is very low
(about 30 m3/s), and the transfer of water from the
Mantaro to the Rímac River is essential for supplying
Lima’s population (about nine million people).
Annual maximum and minimum monthly runoff
(Qmax and Qmin, respectively) complement the mean
annual runoff (Qmean). They are the monthly runoff
values of the months with the highest and lowest
runoff values respectively. The specific runoff of the
rivers was computed in order to eliminate the relation-
ship between runoff and basin size.
Seasonal variation coefficients (SVC) were cal-
culated using mean monthly rainfall and runoff and
their standard deviation. Likewise, multi-annual vari-
ation coefficients (MAVC) were computed using
annual rainfall and runoff values.
The trend in the hydrological series was evalu-
ated using the non-parametric Mann-Kendall test that
is based on range probability of the data occurrence
order (Mann 1945, Kendall 1975) and is considered
an excellent tool (Zhang et al. 2001, Burn and Elnur
2002).
In addition, the Pettitt non-parametric test for
mean-change measurement in temporal hydrological
annual series was applied (Pettitt 1979). Considered
Basin-scale analysis of rainfall and runoff in Peru (1969–2004): Pacific, Titicaca and Amazonas drainages 631
as one of the most complete tests for identifying
mean-change (Kundzewicz and Robson 2004), it is
based on changes in the range of the values subdi-
vided into sub-series. The significance level (α) is the
probability that a test detects trend or other change
when none is present (Robson et al. 2000). In this
study, the null hypothesis in statistical tests is rejected
using α = 0.05.
Several regional climatic indexes, provided by
the Climatic Prediction Centre of the National
Oceanic and Atmospheric Administration (CPC-
NOAA, http://www.cdc.noaa.gov/), were used to
understand the time variability of rainfall and runoff.
The Southern Oscillation Index (SOI) is the standard-
ized pressure difference between Tahiti and Darwin.
Monthly SSTs are provided for the northern tropical
Atlantic (NATL, 5–20◦N, 60–30◦W) and the southern
tropical Atlantic (SATL, 0–20◦S, 30◦W–10◦E). The
standardized SST difference between the NATL and
SATL was computed to determine the SST gradient.
Finally, the management of rainfall and runoff
databases, as well as the calculation of the aver-
age rainfall in the basins, was carried out using
the HYDRACCESS software, developed within the
framework of the HyBAm project (Vauchel 2005).
SPATIAL AND SEASONAL VARIABILITY OF
RUNOFF AND RAINFALL
Mean annual rainfall values (1969–2004) in 167 sta-
tions, interpolated using the kriging method, are
displayed in Fig. 2. They feature a strong contrast
between the rainy eastern lowlands in the Amazon
basin (more than 1500 mm/year) and much drier con-
ditions in the Andes and almost no rainfall in the
basins of the Pacific coast, except in the extreme
north. When integrating rainfall by basin (Table 2),
annual values vary from about 144 mm year-1 in the
southernmost Pacific coast basin, to 1900 mm in the
Tabatinga basin.
Along the Pacific coast, mean rainfall per basin
increases regularly as latitude diminishes (i.e. from
south to north), from 144 mm/year (PQ-29) to nearly
1000 mm/year in the north (Table 2 and Fig. 3(a)).
The specific runoff tends to increase consis-
tently towards the north (Fig. 3(b)), ranging from
less than 1 L s-1 km-2 in the south to 24 L s-1 km-2
in the northernmost Tumbes basin (Table 1); but
the specific runoff increase is not as regular as the
rainfall increase. Indeed, it seems that the relation
between specific runoff and latitude is organized
within some sets of stations. These groups of stations,
from PQ-1 to PQ-3, from PQ-4 to PQ-9, from PQ-
10 to PQ-20 and finally from PQ-21 to PQ-29 are
lined up in Fig. 3(b).
Along the Pacific coast, there is no clear rela-
tionship between annual in situ rainfall and altitude
(Fig. 4(a)). Near null or very low rainfall values are
often observed near sea-level. Above the inversion
layer, rainfall is generally more abundant (Dollfus
1964, Johnson 1976) but very little rainfall may
also be recorded at altitude, probably due to the lee
wind exposure of the stations. The highest rainfall,
up to 2000–2500 mm/year, is observed at between
1000 and 3000 m, but very high rainfall variabil-
ity characterizes this altitudinal range. Finally, above
3000 m, rainfall increases from less than 500 to
750 mm/year.
In Td annual rainfall varies from 400 to 800 mm
per year below 4000 m, and does not exceed 600 mm
above this altitude (Fig. 4(b)). As a consequence of
the moderate rainfall, specific runoff is about 5 L s-1
km-2 (Table 1).
In the Ad, in contrast, in situ rainfall is abundant
and, as a consequence, specific runoff is greater than
40 L s-1 km-2 in Tabatinga and Tamshiyacu (Table 1).
Rainfall is very high near Equator (>2500 mm/year)
and local maximum and minimum rainfall are notice-
able in Fig. 2. In particular, very high annual rain-
fall (>6000 mm/year) is recorded in regions where
the concave form of the Andes favours convergence
(Borja region), in valleys open to the north and ori-
ented toward the humid northwestern winds (High
Ucayali and Pachitea valleys), and along massive
relief with a northward exposure. They contrast with
less rainy regions such as the high Huallaga valley
that is located on a lee side and is consequently pro-
tected from moist air advection. In the Amazon basin,
local rainfall tends to increase as altitude diminishes
(Fig. 4(c)). Abundant rainfall of ∼2500 mm/year is
related to the moist warm air and to the release of high
quantities of water vapour over the first eastern slope
of the Andes. In contrast, annual rainfall is often less
than 1000 mm/year above 500 m elevation. Similar
situations regarding rainfall–topography relationships
have been described in the Amazon basin as a whole
by Johnson (1976), Figueroa and Nobre (1990) and
Espinoza et al. (2009b), in Ecuador, Colombia and
Venezuela by Laraque et al. (2007), Poveda (2004)
and Pulwarty et al. (1992), respectively, and in
Bolivia by Guyot (1993) and Ronchail and Gallaire
(2006).
Monthly rainfall per basin and runoff values,
expressed as a percentage of the hydrological annual
632 Waldo Sven Lavado Casimiro et al.
Fig. 2 Interpolated mean annual rainfall values (mm) for the 1970–2004 period, over the (a) Pacific (Pd), Amazonas (Ad)
and Titicaca (Td) drainages, using in situ data.
values, are presented in Fig. 5 for all the gauging
stations. Along the Pacific coast (Pd) and in Td, there
is a pronounced annual cycle in rainfall, with a rainy
season during the austral summer and a dry season
in winter. On the coast, the peak of the rainy season
changes with latitude. South of 12◦S, from PQ-14 to
PQ-29, there is a January–February peak consistent
with the occurrence of a rainfall peak in the tropi-
cal Andes that provides most of the water in these
basins. From PQ-14 to the north of the Peruvian coast
(near the Equator) the rainfall peak is more acute
and occurs in March, i.e. during the equinox period.
Moreover, the dry season diminishes from south to
north; the period without rainfall or with very little
rainfall begins as soon as April in the south and in
June near the Equator. The strong seasonal variability
is also illustrated by the seasonal coefficient of vari-
ation (SCV) values. They are always higher than 0.7
(Table 2) and decrease from the southern to the north-
ern Pacific coast. SCV values as high as 1.5 are
computed for the southern stations. Evidently, there
is a strong seasonality in discharge in the coastal
basins (Table 2). This is enhanced by the fact that
some small rivers, in the north as well as in the south
of the Pacific coast, are intermittent and are char-
acterized by null or near null low-level flow in the
austral winter. Moreover, it is noticeable that in some
basins the seasonal discharge variability is higher
than rainfall variability, while the contrary is true in
other basins where runoff is regulated by man (for
Basin-scale analysis of rainfall and runoff in Peru (1969–2004): Pacific, Titicaca and Amazonas drainages 633
-18 -16 -14 -12 -10 -8 -6 -4 -2
0
5
10
15
20
25
PQ-1
PQ-2
PQ-3
PQ-4
PQ-5
PQ-6
PQ-7
PQ-8PQ-9
PQ-10
PQ-11PQ-12PQ-13
PQ-14
PQ-15
PQ-16
PQ-17
PQ-18
PQ-19
PQ-20
PQ-21
PQ-22
PQ-23
PQ-24
PQ-25PQ-26
PQ-27PQ-28
PQ-29
Specific Runoff
Latitude (°S)
(L
 s
-1
 K
m
2 )
-18 -16 -14 -12 -10 -8 -6 -4 -2
300
600
900 PQ-1
PQ-2
PQ-3
PQ-4PQ-5
PQ-6
PQ-7
PQ-8
PQ-9
PQ-10
PQ-11PQ-12
PQ-13PQ-14
PQ-15
PQ-16PQ-17
PQ-18
PQ-19
PQ-20
PQ-212
PQ-23PQ-24
PQ-25PQ-26
PQ-27
PQ-28PQ-29
Rainfall(a)
(b)
Latitude (°S)
(m
m
 y
ea
r-1
)
Fig. 3 Relationships between latitude and: (a) mean multi-annual rainfall per basin (mm), and (b) specific runoff
(L s-1 km2), in the Pacific drainage. Labels and locations of the stations are described in Fig. 1 and Table 1.
0 1000 2000 3000 4000 5000 6000
0
1000
2000
3000
Pacific
(a)
(b)
(c)
Elevation (m a. s.l.)
R
ai
nf
al
l (
m
m
)
3800 3900 4000 4100 4200 4300 4400 4500 4600 4700 4800
0
500
1000
Titicaca
Elevation (m a.s.l.)
R
ai
nf
al
l (
m
m
)
0 500 1000 1500 2000 2500 3000 3500 4000 4500 5000
0
2000
4000
6000
8000
Amazonas
Elevation (m a.s.l.)
R
ai
nf
al
l (
m
m
)
Fig. 4 Relationships between altitude (m a.s.l.) and mean multi-annual in situ rainfall (mm) during the 1970–2004 period
in: (a) the Pacific drainage, Pd; (b) the Titicaca drainage, Td; and (c) the Amazonas drainage, Ad.
634 Waldo Sven Lavado Casimiro et al.
S O N D J F M A M J J A
0
5
10
15
20
25
30
35
40
PQ-1
S O N D J F M A M J J A
0
5
10
15
20
25
30
35
40
PQ-2
S O N D J F M A M J J A
0
5
10
15
20
25
30
35
40
PQ-3
S O N D J F M A M J J A
0
5
10
15
20
25
30
35
40
PQ-4
S O N D J F M A M J J A
0
5
10
15
20
25
30
35
40
PQ-5
S O N D J F M A M J J A
0
5
10
15
20
25
30
35
40
PQ-6
S O N D J F M A M J J A
0
5
10
15
20
25
30
35
40
PQ-7
S O N D J F M A M J J A
0
5
10
15
20
25
30
35
40
PQ-8
S O N D J F M A M J J A
0
5
10
15
20
25
30
35
40
PQ-9
S O N D J F M A M J J A
0
5
10
15
20
25
30
35
40
PQ-10
S O N D J F M A M J J A
0
5
10
15
20
25
30
35
40
PQ-11
S O N D J F M A M J J A
0
5
10
15
20
25
30
35
40
PQ-12
S O N D J F M A M J J A
0
5
10
15
20
25
30
35
40
PQ-13
S O N D J F M A M J J A
0
5
10
15
20
25
30
35
40
PQ-14
S O N D J F M A M J J A
0
5
10
15
20
25
30
35
40
PQ-15
S O N D J F M A M J J A
0
5
10
15
20
25
30
35
40
PQ-16
S O N D J F M A M J J A
0
5
10
15
20
25
30
35
40
PQ-17
S O N D J F M A M J J A
0
5
10
15
20
25
30
35
40
PQ-18
S O N D J F M A M J J A
0
5
10
15
20
25
30
35
40
PQ-19
S O N D J F M A M J J A
0
5
10
15
20
25
30
35
40
PQ-20
S O N D J F M A M J J A
0
5
10
15
20
25
30
35
40
PQ-21
S O N D J F M A M J J A
0
5
10
15
20
25
30
35
40
PQ-22
S O N D J F M A M J J A
0
5
10
15
20
25
30
35
40
PQ-23
S O N D J F M A M J J A
0
5
10
15
20
25
30
35
40
PQ-24
S O N D J F M A M J J A
0
5
10
15
20
25
30
35
40
PQ-25
S O N D J F M A M J J A
0
5
10
15
20
25
30
35
40
PQ-26
S O N D J F M A M J J A
0
5
10
15
20
25
30
35
40
PQ-27
S O N D J F M A M J J A
0
5
10
15
20
25
30
35
40
PQ-28
S O N D J F M A M J J A
0
5
10
15
20
25
30
35
40
PQ-29
S O N D J F M A M J J A
0
5
10
15
20
25
30
35
40
TQ-1
S O N D J F M A M J J A
0
5
10
15
20
25
30
35
40
TQ-2
S O N D J F M A M J J A
0
5
10
15
20
25
30
35
40
TQ-3
S O N D J F M A M J J A
0
5
10
15
20
25
30
35
40
AQ-1
S O N D J F M A M J J A
0
5
10
15
20
25
30
35
40
AQ-2
Fig. 5 Mean monthly per basin rainfall and runoff in percentage (ratio between monthly and mean annual values for runoff
and ratio between monthly and total annual values for rainfall) in the selected basins in Peru (see Fig. 1(a) and Table 1). The
solid lines represent rainfall and the dashed lines runoff.
instance from PQ-11 to PQ-15, around Lima, and in
the southernmost basins).
On the tropical Altiplano (Td), there is a sharp
contrast between a rainy season in the austral summer,
with a peak in January, and a very dry season from
May to August. Consequently, the seasonal variability
of rainfall and discharge is important and SCV values
are about 1 (Table 2). In contrast, in the Ad, rain-
fall is abundant all year round, with a small summer
maximum during the South American monsoon and
a minimum in August. The seasonal rainfall and dis-
charge coefficients of variation are much lower than
in Pd and Td, i.e. 0.2 for rainfall per basin and 0.3 for
discharge (Table 2).
In the small Pd basins there is an immediate
runoff response to rainfall, with rainfall and discharge
peaks in March and low values in the austral win-
ter (Fig. 5). But, towards the south there may be
a month’s lag between the maximum rainfall and
the high flow runoff. In the Sama basin PQ-28, for
instance, which is well-known for its dryness and
very high solar radiation, the lag may be explained
by the delay between rainfall, the soil moistening
after the dry season and runoff. In Td, there is a lag
of about one month between rainfall (January) and
runoff (February–March) peaks. As the sizes of the
Amazonian basins are huge, there are two-month lags
between peaks in rainfall (March) and runoff (May),
as in Tabatinga gauging station (AQ-1).
INTER-ANNUAL VARIABILITY OF
RAINFALL AND DISCHARGE
Standardized 1970–2005 annual runoff and rainfall
per basin are plotted for some stations (Fig. 6), show-
ing that rainfall and runoff present similar variability.
This is confirmed by the correlation between rainfall
and runoff (Table 2). The coefficients of correlation
range between 0.4 and 0.96; that means that rain-
fall explains 16% of discharge variability in some
basins and 92% in others. The correlation is low
in some coastal basins (PQ-14, PQ-15, from PQ-
18 to PQ-23, PQ-27) and not significant in the Santa
basin (PQ-10). This may be due to various reasons,
natural or anthropogenic. The availability of under-
ground water may sustain low flow runoff when
rainfall is missing (Mark et al. 2010); its absence may
emphasize the seasonal and inter-annual variability.
Glacier-ice melting in upstream valleys, particularly
important at the beginning of the rainy season when
the glaciers are “dirty” and the albedo is very low
(Wagnon et al. 1999, 2001, Sicart 2002) may also
Basin-scale analysis of rainfall and runoff in Peru (1969–2004): Pacific, Titicaca and Amazonas drainages 635
1970 1975 1980 1985 1990 1995 2000 2005
-10
-7.5
-5
-2.5
0
2.5
5
7.5
10
Tumbes (PQ-1)(a) (b)
(c) (d)
(e) (f)
1970 1975 1980 1985 1990 1995 2000 2005
-10
-7.5
-5
-2.5
0
2.5
5
7.5
10
Santa (PQ-10)
1970 1975 1980 1985 1990 1995 2000 2005
-10
-7.5
-5
-2.5
0
2.5
5
7.5
10
Cañete (PQ-17)
1970 1975 1980 1985 1990 1995 2000 2005
-10
-7.5
-5
-2.5
0
2.5
5
7.5
10
Majes (PQ-24)
1970 1975 1980 1985 1990 1995 2000 2005
-10
-7.5
-5
-2.5
0
2.5
5
7.5
10
Ramis (TQ-1)
1970 1975 1980 1985 1990 1995 2000 2005
-10
-7.5
-5
-2.5
0
2.5
5
7.5
10
Tabatinga (AQ-1)
Fig. 6 Annual hydrological series (September–August) for: (a) the Tumbes River (PQ-1), (b) the Santa basin (PQ–10),
(c) the Cañete basin (PQ-17), (d) the Majes basin ( PQ-24), (e) the Ramis basin (TQ-1) and (f) the Tabatinga basin (AQ-1).
Black lines represent: mean annual runoff; grey lines: total annual per basin rainfall; dotted lines: maximum runoff; and
dashed lines: minimum runoff. Values are standardized and are corrected by coefficients in order to avoid confusion between
the different lines. The coefficients are 4 for maximum runoff, 0 for mean runoff and –4 for minimum runoff.
sustain the runoff. This may explain some discon-
nection between rainfall and discharge in the Santa
River (PQ-10) located downstream of the Cordillera
Blanca glaciers (Fig. 6). Indeed, Mark et al. (2005)
estimate that nearly 40% of the upper Rio Santa dis-
charge at Huallanca (which comprises 40% of the Rio
Santa basin at Puente Carretera—PQ-10) is glacier
melt. The nature of the link between rainfall and
discharge can also be driven by barrages or water
abstraction/deviation (for irrigation, hydroelectricity
or city water supply), as in the case of the Rímac basin
(PQ-15) that provides water towards the city of Lima
and of the Santa basin (PQ-10) that provides water for
hydroelectricity and irrigation purposes.
Figure 7 presents the gross runoff and rain-
fall per basin and points out a strong inter-annual
variability along the coast that culminates in the
Piura River (PQ-3) where the rainfall multi-annual
variation coefficient (MAVC) reaches 0.8 and where
dischargeMAVC surpasses 1 and attains 3.3 for Qmin
(Table 2). The highest values are often observed in
the north, in association with ENSO events that some-
times quintuple annual rainfall. Dramatic rainfall and
runoff anomalies were observed during the last two
extreme ENSO events (1982–1983 and 1997–1998)
(Figs 6(a) and 7(a)). However, discharge variabil-
ity is not spatially even: some basins with a very
strong inter-annual variability, such as Piura (PQ-3),
are next to others that display a very low variabil-
ity (Chira—PQ-2 or Chancay-Lambayeque—PQ-4).
These differences are probably due to local ground-
water and/or water exploitation conditions. On the
southern Pacific coast, high runoff MAVC are con-
sistent with rainfall variability and may be accen-
tuated by the lack of outflow during some very
dry years, i.e. by the intermittency of the rivers
(Fig. 7(b)).
In Td and Ad, correlation between rainfall and
runoff is generally high, as runoff does not stop
seasonally and is not perturbed by anthropogenic
activities (Table 2). In Tamshiyacu (AQ-2), the rather
weak correlation (r = 0.6) between rainfall and runoff
may be related to the poor availability of raingauge
data in the rainy Marañon sub-basin of the Amazon
(see Fig. 1(c)). A weak inter-annual variability in the
Ad gives very low MAVC, about 0.1 for rainfall,
while they are a little higher in the Td (Table 2 and
see Tabatinga and Ramis in Figs 7(d) and (c)).
636 Waldo Sven Lavado Casimiro et al.
1970 1980 1990 2000
0
1000
2000
3000
4000
Tabatinga (AQ-1)
1970 1980 1990 2000
0
1000
2000
3000
4000
Pte. Ramis (TQ-1)
1970 1980 1990 2000
0
1000
2000
3000
4000
Pte. Ñacara (PQ-3)
1970 1980 1990 2000
0
1000
2000
3000
4000
 La Pascana (PQ-26)
(a)
(c) (d)
(b)
Fig. 7 Annual hydrological series (September–August): (a) Pte Ñacara (PQ-3); and (b) La Pascana (PQ-26) in the Pacific
drainage; (c) Pte Ramis (TQ-1) in the Titicaca drainage; and (d) Tabatinga (AQ-1) in the Amazonas drainage. Bars represent:
total annual per basin rainfall; solid line: mean annual runoff; dotted lines: maximum runoff; and dashed lines: minimum
runoff. Values are in mm.
In summary, the Pd, shows greater rainfall and
runoff variation than Td and Ad, at the seasonal
and inter-annual time scales. This is particularly vis-
ible in Figs 8(a) and (c) where MAVC of runoff
and rainfall, respectively, are plotted against sea-
sonal coefficients (SVC). The Titicaca and Amazon
basins are located near the origin of the graph, while
the Pacific basins are generally located higher in
the graph, occupying a wide range of values and
demonstrating a large variety of natural and anthro-
pogenic situations. Figures 8(b) and (d) show the ratio
MAVC/SVC in runoff and rainfall in Pd vs latitude.
As already mentioned, a clear difference is evident
between northern and southern stations, the northern
Pd basins showing the highest inter-annual variation
in rainfall and runoff, when compared to seasonal
variation, due to the strong hydrological impacts of
extreme El Niño events on the northern coast. Thus,
during the last strong El Niño events (1982–1983 and
1997–1998) rainfall and runoff changes (related to the
1969–2004 mean) in the northern Pd were approxi-
mately +250% (PQ-1, PQ-2, PQ-3 and PQ-4, north-
ward of 6.5◦S.).
Another set of basins, in the southernmost coastal
region, experiences a strong inter-annual runoff vari-
ability, which is not always observed in rainfall. This
may be due to the use of abundant rainfall gauges in
the Andes to compute rainfall over the basins when
compared to the number of rainfall gauges along the
coast. Yet, we know that rainfall variability is lower in
the Andes than on the coast.
RUNOFF AND RAINFALL CHANGES AND
TRENDS
The long-term variability of rainfall and runoff is
analysed through break and trend analysis. The years
of significant mean changes in rainfall and runoff in
the Peruvian basins are presented in parentheses in
Table 3, together with information about trend in the
series.
The main result is a change in rainfall and runoff
in the mid-1980s in the two Amazonian basins, with
rainfall and runoff decreases afterwards that have
already been described by Espinoza et al. (2009a).
The decrease is particularly important in Qmin since
1987, principally in Tamshiyacu (from 18 911 m3 s-1
in 1969 to 15 390 m3 s-1 in 2006). In this station, the
absence of significant rainfall decrease may be due
to uncertainties about rainfall estimation related to
the low rainfall data availability in the Marañón sub-
basin (Fig. 2). In Tabatinga, the Qmin decrease (from
Basin-scale analysis of rainfall and runoff in Peru (1969–2004): Pacific, Titicaca and Amazonas drainages 637
0.2 0.4 0.6 0.8 1 1.2 1.4 1.6
0
0.5
1
1.5
AQ-1AQ-2
PQ-1PQ-2
PQ-3
PQ-4
PQ-5
PQ-6
PQ-7
PQ-8
PQ-9
PQ-10
PQ-11
PQ-12
PQ-13PQ-14
PQ-15
PQ-16PQ-17
PQ-18
PQ-19
PQ-20
PQ-21PQ-22
PQ-23
PQ-24
-25
PQ-26
PQ-27
PQ-28
PQ-29
TQ-1
TQ-2
TQ-3
SVC
M
A
V
C
Mean Runoff
-18 -16 -14 -12 -10 -8 -6 -4 -2
0.4
0.5
0.6
0.7
0.8
0.9
1
1.1
1.2
1.3
PQ-1
PQ-2
PQ-3
PQ-4
PQ-5
PQ-6
PQ-7
PQ-8
PQ-9
PQ-10
PQ-11
PQ-12PQ-13
PQ-14
PQ-15
PQ-16
PQ-17PQ-18
PQ-19
PQ-20
PQ-21
PQ-22
PQ-23
PQ-24
PQ-25
PQ-26
PQ-27
PQ-28
PQ-29
Latitude (°S)
Mean Runoff
0.2 0.4 0.6 0.8 1 1.2 1.4 1.6
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
AQ-1AQ-2-5
AQ-4
AQ-3
PQ-1
PQ-2
PQ-3
PQ-4
PQ-5
PQ-6
PQ-7PQ-8PQ-9
PQ-10
PQ-11
PQ-12
PQ-13PQ-14
PQ-15
PQ-16
PQ-17
PQ-18
PQ-19
PQ-20
PQ-21
PQ-22
-23PQ-24PQ-25
- 6
PQ-27
PQ-28
PQ-29
TQ-1TQ-2
TQ-3
SVC
M
A
V
C
Rainfall
-18 -16 -14 -12 -10 -8 -6 -4 -2
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
1.1
1.2
PQ-1PQ-2
PQ-3
PQ-4
PQ-5
PQ-6
PQ-7
PQ-8
PQ-9
PQ-10
PQ-11PQ-12PQ-13PQ-14PQ-15
PQ-16
PQ-17PQ-18
PQ-19
PQ-20
PQ-21
PQ-22PQ-23
PQ-24PQ-25
PQ-26PQ-27
PQ-28
PQ-29
Latitude (°S)
M
A
V
C
 S
V
C
-1
M
A
V
C
 S
V
C
-1
Rainfall
(a) (b)
(c) (d)
Fig. 8 Relationships between multi-annual variation coefficients (MAVC) and seasonal variation coefficients (SVC) for:
(a) runoff and (c) per basin rainfall, in Peru; and the MAVC/SVC ratio as a function of latitude for stations located in
the Pacific drainage: (b) runoff and (d) per basin rainfall. Black circle: Pacific drainage stations; grey diamond: Amazon
drainage stations; black upward triangle: Titicaca drainage stations.
25 316 m3 s-1 in 1969 to 22 126 m3 s-1 in 2006) is
associated with a rainfall diminution.
In the Td, in contrast, there is a significant
increase in minimum runoff, from 6 m3 s-1 in 1969 to
13 m3 s-1 in 2006, in the Ramis basin, with a change
in 1986 that is not related to a change in annual rain-
fall. The analysis of seasonal rainfall, in particular
during the end of the rainy season and during the
low flow season, would be useful to understand this
feature.
Along the Pd, half of the stations present changes
and trends, mainly in minimum runoff. The signals
differ from one station to another, occur at very dif-
ferent dates from the 1970s until the late 1990s and
are not related to change in annual rainfall. As there
is no uniform regional signal, we can hypothesize
that the atmospheric circulation and rainfall are not
the cause of these changes. They may be related to
constructions dedicated to sustaining low-flow “dam
intakes” when the change is positive, and to water
withdrawal when the change is negative. Positive
trends are found in the Chillon (PQ-14), Rímac (PQ-
15), Yauca (PQ-22) and Sama (PQ-28). The runoff
of the Chillon and Rímac rivers is mostly diverted to
urban use for the city of Lima that, according to the
2007 census from the National Institute of Statistic
and Informatics of Peru, INEI (http://www.inei.gob.
pe), gathers 30.8% of the total Peruvian population
(about 28 million). Thus, the increase in minimum
runoff is due to runoff control in the high rainfall
zones of these basins, which involve many reser-
voir systems. Negative trends are observed in the
rivers Viru (PQ-9), Pativilca (PQ-11), Huaura (PQ-
12), Acari (PQ-21) and Ocoña (PQ-23).
There is no trend in the big Santa River (PQ-
10); on the one hand, its upper basin benefits
from the melting of the Cordillera Blanca Glacier
(Mark and Seltzer 2003, Mark et al. 2005, Pouyaud
et al. 2005) but, on the other hand, its water is
intensely exploited for export agricultural production
and hydro-electricity.
RELATIONSHIPS BETWEEN
OCEAN–ATMOSPHERE INDICES AND
PERUVIAN HYDROLOGY
In order to find explanations for the inter-annual
and long-term variability of runoff and rainfall per
638 Waldo Sven Lavado Casimiro et al.
Table 3 Changes and trends in runoff and per basin rain-
fall. Years of change are in parentheses when the Pettit
test is significant at the 95% level. Statistical trends are
computed using the Mann-Kendall test for runoff (max-
imum, mean and minimum) and rainfall annual series
(1969–2004 period) except for PQ-10 (1969–1999 for
runoff). NEG: significant negative trends (95% confidence
level); POS: significant positive trends (95% confidence
level); nt: no trend; ∗∗ not computed.
Code Trend (YC)
Maximum
runoff
Mean
runoff
Minimum
runoff
Rainfall
PQ-1 nt nt nt nt
PQ-2 nt nt nt nt
PQ-3 nt nt nt (1991) nt
PQ-4 nt nt NEG (1990) nt
PQ-5 nt nt NEG nt
PQ-6 nt nt nt nt (1991)
PQ-7 nt nt nt nt
PQ-8 nt nt nt nt
PQ-9 nt nt NEG (1976) nt
PQ-10 nt (1989) nt nt nt
PQ-11 nt nt NEG (1980) nt
PQ-12 nt nt NEG (1988) NEG
PQ-13 nt nt nt (1988) nt
PQ-14 nt nt POS (1994) nt
PQ-15 nt nt POS (1994) NEG
PQ-16 NEG nt Nt nt (1992)
PQ-17 nt nt Nt nt
PQ-18 nt (1976) nt Nt nt
PQ-19 nt (1988) nt POS (1991) nt (1992)
PQ-20 NEG nt ∗∗ nt
PQ-21 nt nt NEG (1982) POS (1992)
PQ-22 nt nt POS (1986) nt
PQ-23 nt nt NEG (1989) nt
PQ-24 nt nt nt nt
PQ-25 nt nt nt nt
PQ-26 nt nt nt nt
PQ-27 nt nt nt nt
PQ-28 nt nt POS (1997) nt
PQ-29 nt nt nt nt
TQ-1 nt nt POS (1986) nt
TQ-2 nt nt nt nt
TQ-3 nt nt nt nt
AQ-1 NEG
(1984)
NEG
(1984)
NEG (1987) NEG (1984)
AQ-2 NEG
(1984)
NEG
(1987)
NEG (1987) nt (1978)
AQ-3 nt
AQ-4 nt
AQ-5 nt
basin, the hydrological series have been correlated
to ocean–atmosphere indices, such as the Southern
Oscillation Index (SOI), and Atlantic indexes con-
structed using sea-surface temperature (SST) in the
southern and northern tropical Atlantic Ocean. These
indexes were chosen because coastal Peru is a key
region for ENSO phenomena (Francou and Pizarro
1986, Aceituno 1988, Tapley and Waylen 1990,
Kane 1999, among others), and because the tropical
Atlantic SST controls the location of atmospheric
ascendance and subsidence over the Atlantic Ocean
and South America, the strength of the trade winds
and water vapour advection from the Atlantic to
the South America tropics (Marengo 2004, Garreaud
et al. 2009). Moreover, Espinoza et al. (2009b)
showed that the SSTs in the northern tropical Atlantic
are significantly correlated with discharge in the west-
ern Amazonian rivers.
Correlations values using the Mann-Kendall test
are reported in Table 4. They are often low, explain-
ing, at best, 25–30% of the hydrological series vari-
ability. Correlations between minimum runoff and
ocean–atmosphere indexes were not considered as
reliable, as they can be influenced by human activi-
ties, ice melting and the intermittency of the rivers to
a much greater extent than Qmean and Qmax,
In the north of the Pd, the expected negative rela-
tionships between ENSO and rainfall or runoff are not
observed, as heavy rainfall is not observed during all
El Niño events but mainly during extreme episodes
(1982/83 and 1997/98, see Fig. 7). In the south, the
ENSO signal is similar to that observed usually in the
Andes, as most of the rainfall in the basins comes
from the Andes. The same positive signal is observed
in Qmean and Qmax values in all the southern basins,
indicating that runoff is lower than usual during El
Niño. Moreover, rainfall, Qmean and Qmax in all
the southern basins (from PQ-11 for rainfall and PQ-
16 for runoff, to PQ-29) are negatively related to the
SST gradient in the tropical Atlantic (NATL–SATL).
This means that runoff is higher when the south-
ern tropical Atlantic is warmer than usual and/or the
northern tropical Atlantic is cooler than usual. Once
again, this relationship may represent the link between
Atlantic SST and hydrology in the Andes as the water
comes from this region. A recent study by Herreros
et al. (2009) is in agreement with this result.
In the Td, rainfall is usually described as lower
during El Niño, but this signal is weak and signifi-
cant only in the Ramis basin rainfall (TQ-1) and in the
mean and maximum discharge of the Huancane River
(TQ-2). As observed on the southern coast, there is a
negative relationship between NATL, or the Atlantic
SST gradient, and hydrological series, but it is rather
weak and not all the basins or all the variables are
affected.
In Tabatinga, in the Ad, rainfall and runoff are
positively related to SOI, with less rainfall and lower
mean and maximum runoff during El Niño. Thus,
in the Ad, there is a negative signal between NATL
Basin-scale analysis of rainfall and runoff in Peru (1969–2004): Pacific, Titicaca and Amazonas drainages 639
Table 4 Mann-Kendall coefficients between hydrological time series (maximum, mean and minimum runoff and per basin
rainfall) and large-scale circulation indexes (SOI: Southern Oscillation Index, NATL: North Atlantic SST, N-S: NATL–
SATL SSTs). Significant values (95% confidence level) are underlined (negative) or bold (positive). ∗∗ not computed.
Code Maximum runoff Mean runoff Minimum runoff Rainfall
SOI NATL N-S SOI NATL N-S SOI NATL N-S SOI NATL N-S
PQ-1 0.19 −0.31 −0.12 0.12 −0.42 −0.22 0.08 −0.46 −0.16 0.29 −0.26 −0.13
PQ−2 −0.10 −0.05 −0.01 −0.08 −0.12 −0.04 −0.07 0.03 −0.05 −0.06 −0.16 −0.10
PQ−3 0.11 −0.19 −0.22 0.02 −0.29 −0.30 0.13 −0.40 −0.30 −0.06 −0.16 −0.10
PQ-4 −0.06 −0.11 −0.08 −0.05 −0.12 −0.13 −0.09 0.10 −0.04 −0.13 −0.02 −0.01
PQ-5 0.08 −0.13 −0.18 0.13 −0.20 −0.25 0.27 −0.43 −0.36 0.13 0.03 −0.09
PQ-6 −0.07 −0.16 −0.11 0.05 −0.20 −0.16 0.38 −0.40 −0.30 0.01 −0.07 −0.12
PQ-7 0.03 0.00 −0.09 0.10 −0.12 −0.23 0.21 −0.34 −0.27 −0.09 0.08 0.01
PQ-8 0.05 −0.10 −0.09 0.09 −0.15 −0.21 0.39 −0.40 −0.34 0.10 −0.10 −0.11
PQ-9 0.12 −0.09 −0.14 0.11 −0.10 −0.19 0.38 −0.29 −0.44 0.09 −0.04 −0.12
PQ-10 0.10 −0.05 −0.09 0.11 −0.06 −0.12 0.18 −0.30 −0.27 0.19 −0.17 −0.21
PQ-11 0.12 −0.01 −0.14 0.03 0.02 −0.21 0.08 −0.04 −0.25 0.17 −0.22 −0.33
PQ-12 0.07 −0.11 −0.19 0.15 −0.21 −0.21 0.24 −0.44 −0.10 0.22 −0.21 −0.28
PQ-13 0.01 −0.25 −0.19 0.15 −0.31 −0.31 0.20 −0.46 −0.31 0.17 −0.35 −0.35
PQ-14 0.04 −0.22 −0.19 0.11 −0.18 −0.26 0.13 −0.20 −0.19 0.14 −0.16 −0.32
PQ-15 0.08 −0.17 −0.13 0.05 −0.12 −0.17 0.00 0.10 0.04 0.18 −0.26 −0.33
PQ-16 0.09 −0.16 −0.09 0.02 −0.06 −0.16 −0.01 0.23 0.05 0.20 −0.29 −0.32
PQ-17 0.25 −0.50 −0.32 0.23 −0.43 −0.46 0.29 −0.24 −0.32 0.21 −0.02 −0.25
PQ-18 0.13 −0.24 −0.23 0.11 −0.20 −0.22 0.00 −0.22 −0.03 0.19 −0.15 −0.41
PQ-19 0.33 −0.46 −0.44 0.28 −0.39 −0.52 0.24 −0.33 −0.43 0.12 −0.24 −0.32
PQ-20 0.18 −0.12 −0.25 0.16 −0.16 −0.35 −0.07 −0.09 −0.18 0.13 −0.01 −0.29
PQ-21 0.25 −0.50 −0.32 0.22 −0.42 −0.45 ∗∗ ∗∗ ∗∗ 0.26 −0.16 −0.34
PQ-22 0.13 −0.24 −0.26 0.15 −0.18 −0.28 0.10 0.01 0.39 0.22 0.03 −0.24
PQ-23 0.10 −0.22 −0.27 0.15 −0.16 −0.32 −0.05 −0.02 −0.25 0.19 −0.17 −0.31
PQ-24 0.36 −0.25 −0.31 0.33 −0.32 −0.38 0.17 −0.24 −0.13 0.19 −0.17 −0.41
PQ-25 0.33 −0.32 −0.31 0.35 −0.30 −0.45 0.12 −0.08 −0.02 0.30 −0.14 −0.34
PQ-26 0.33 −0.26 −0.29 0.32 −0.25 −0.30 0.05 −0.10 0.10 0.29 −0.19 −0.33
PQ-27 0.34 −0.22 −0.27 0.35 −0.24 −0.29 −0.03 −0.26 0.06 0.33 −0.27 −0.27
PQ-28 0.10 −0.27 −0.35 0.14 −0.19 −0.19 −0.01 0.08 0.06 0.34 −0.33 −0.25
PQ-29 0.18 −0.37 −0.30 0.25 −0.44 −0.32 0.00 −0.05 −0.15 0.28 −0.29 −0.31
TQ-1 0.17 −0.21 −0.26 0.24 −0.26 −0.25 0.17 −0.21 −0.09 0.30 −0.27 −0.29
TQ-2 0.31 −0.21 −0.10 0.27 −0.29 −0.21 0.17 −0.06 −0.28 0.19 −0.17 −0.20
TQ-3 0.19 −0.24 −0.05 0.16 −0.26 −0.18 −0.03 −0.19 −0.15 0.05 −0.04 −0.05
AQ-1 0.36 −0.30 −0.20 0.39 −0.33 −0.20 −0.10 −0.18 −0.21 0.38 −0.33 −0.28
AQ-2 0.23 −0.35 −0.15 0.13 −0.45 −0.23 0.08 −0.54 −0.20 0.13 −0.18 −0.12
AQ-3 0.23 −0.21 −0.34
AQ-4 0.19 −0.11 −0.13
AQ-5 0.23 −0.12 −0.09
and rainfall or runoff, indicating lower values when
the northern Atlantic SST is warmer than usual.
These signals are rather similar to those displayed
in the Andes and in the southern Pacific coast. This
could have two explanations. First, the associations
between ocean–atmosphere indicators are the same
across the three Peruvian drainages. Secondly, as
many raingauge stations are located in the Andes
and as large parts of the Amazonas basins are in the
Andes, the Andean signal dominates the Amazonian
hydrology. The relationship between in situ rainfall
and the SST in the tropical Atlantic confirms the sec-
ond hypothesis (Fig. 9(d)). Finally, the lack of rainfall
in the Amazonas basins when the SST in the northern
tropical Atlantic is warm may explain the downward
trend in hydrology described in the section above on
Runoff and Rainfall Changes and Trends; indeed, a
positive trend in the NATL SST is observed between
1970 and 2005 (Fig. 9(b)). These results concur with
the results of Espinoza et al. (2009a).
CONCLUSION
For the first time, an analysis of rainfall and runoff
mean conditions and variability has been conducted,
at the basin scale, over the three principal Peruvian
drainages: Pacific, Lake Titicaca and Amazonas, for
the 1969–2004 period. This work takes advantage
of data gathered in 34 basins by SENAMHI as
part of the Hydrogeodynamics of the Amazon Basin
640 Waldo Sven Lavado Casimiro et al.
-4
-2
0
2
4
(a) SOI
NATL
NATL-SATL
Anual Rainfall
(d)
(b)
(c)
-2
0
2
4
-2
0
1
-1
2
1975 1980 1985 1990 1995 20001970 2005
-4°
-6°
0.2
0
-0.2
-0.4
-0.6
-0.8
-1
-8°
-10°
PQ-11
-12°
-14°
-16°
-78° -76°
PQ-21
-74° -72° -70°
1970 1975 1980 1985 1990 1995 2000 2005
1970 1975 1980 1985 1990 1995 2000 2005
Fig. 9 Standardized annual values (1969–2004 period) of: (a) the Southern Oscillation Index (SOI), (b) the northern tropical
Atlantic SST (NATL, 5–20◦N, 60–30◦W), and (c) the standardized difference between NATL and the southern tropical
Atlantic SST (SATL, 0–20◦S, 30◦W–10◦E). (d) Interpolated correlation coefficients between in situ rainfall and NATL–
SATL in the Ucayali and Huallaga basins. Stations with significant negative correlation values are represented by black
downward triangles (95% confidence level). Adapted from Lavado (2010).
(HyBAm) project. It is notable that many of the
29 Pacific basins are influenced by anthropogenic
activities, which makes it difficult to understand some
results.
In addition to the well-known contrast between
the dry coastal basins and the wet eastern lowlands,
details are given about in situ and per basin rainfall
distribution in all regions and about their differ-
ent altitude–rainfall relationships. Along the coast,
a strong south–north rainfall gradient prevails and
there is no clear relationship between rainfall and alti-
tude. In the Amazon, a patchwork of rainier and drier
regions prevails, together with a rainfall decrease with
altitude. Over the Titicaca drainage, annual rainfall is
of intermediate values and it diminishes in the highest
stations.
Runoff variability is strong in the coastal basins
at seasonal and inter-annual time scales, because
rainfall variability is high in this part of the country,
and because the basins are small and the rivers often
intermittent. Rainfall and runoff are more regular in
the Andes at the inter-annual time scale, and in the
Amazon at intra- and inter-annual time scales.
Trends and change points are observed in the
runoff data of Amazonian basins where rainfall
and runoff decrease, especially since the mid-1980s
and during the low stage season. Over the Titicaca
drainage, increases in minimum runoff may be asso-
ciated with accelerated melting of the glaciers due to
climate change. Minimum runoff values in almost all
the coastal basins show some change (change points
and trends) during the last 35 years. As they are
not related to rainfall changes and are not spatially
organized, they may be attributed to human activity.
Increases in low stage runoff may concur with glacier
melting in the Santa basin and with constructions ded-
icated to sustaining low flow in the rivers that supply
water to big cities or intensive farming, but further
analysis is necessary to assess this hypothesis.
The analysis of the relationships between hydrol-
ogy and ocean–atmosphere indicators, such as the
Pacific Southern Oscillation Index (SOI) and the sea-
surface temperature (SST) over the tropical Atlantic
Ocean, gives some indications about the origin of
rainfall and runoff variability. A signal that had
not been documented before was found between the
Basin-scale analysis of rainfall and runoff in Peru (1969–2004): Pacific, Titicaca and Amazonas drainages 641
Atlantic SST and the hydrology of the southern Andes
basins and of the coastal and Amazon drainages fed
by the southern Andes. Rainfall and maximum and
mean runoff values decrease when the northern trop-
ical Atlantic Ocean is warmer than usual. However,
a cooler ocean may displace convection towards the
southern tropics and drive strong trade winds and
high water vapour fluxes towards the region, enhanc-
ing rainfall and runoff. The long-term warming of
the northern tropical Atlantic may explain the sig-
nificant downward trend of runoff observed in the
Amazon. As documented before, rainfall and runoff
tend to decrease during El Niño events in the tropi-
cal Andes and, consequently, in the dependent coastal
and Amazonian basins. There is no clear evidence of
increased rainfall and runoff on the northern coast
during El Niño, except for some extreme events
(1982/83 and 1997/98). Further investigations may
clarify and complete these first results.
REFERENCES
Aceituno, P., 1988. On the functioning of the Southern Oscillation
in the South American sector. Part 1: Surface climate. Monthly
Weather Review, 116, 505–524.
Aceituno, P. and Montecinos, A., 1993. Circulation anomalies associ-
ated with dry and wet periods in the South American Altiplano.
Proc. Fourth Int. Conf. On Southern Hemisphere Meteorology,
American Meteorological Society.
ACTO, 2005. A proposal for defining the geographical boundaries
of Amazonia. In: H.D. Eva and O. Huber, eds., Synthesis of
the results from an Expert Consultation Workshop organized by
the European Commission in collaboration with the Amazon
Cooperation Treaty Organization – JRC Ispra, 7–8 June 2005.
Brunet-Moret, Y., 1979. Homogénéisation des précipitations. Cahiers
ORSTOM, Série Hydrologie, 16, 3–4.
Burn, D.H. and Elnur, M.A.H., 2002. Detection of hydrologic trends
and variability. Journal of Hydrology, 255, 107–122.
Chaffaut, I. et al., 1998. Précipitations d’altitude du Nord-Chili, orig-
ine des sources de vapeur et données isotopiques. Bulletin de
l’Institut Française d’Etudes Andines, 27, 367–384.
Deutsch, C.V. and Journel, A.G., 1992. GSLIB Geostatistical
Software Library and User’s Guide. TIC: 224174. New York,
New York: Oxford University Press.
Dollfus, O., 1964. Cambios climáticos cuaternarios en los Andes
Peruanos. Boletin de la Sociedad Geografica. Lima, 83, 65–74
Espinoza, J.C. et al., 2006. La variabilité des débits du Rio Amazonas
au Pérou. In: Climate variability and change—Hydrological
impacts (Proceedings of the Fifth FRIEND World Conference
held at Havana, Cuba, November 2006). Wallingford, UK:
IAHS Press, IAHS Publ. 308, 424–429.
Espinoza, J.C. et al., 2009a. Contrasting regional discharge evolutions
in the Amazon basin (1974–2004. Journal of Hydrology, 375,
297–311.
Espinoza, J.C. et al., 2007. Hydroclimatic variability in the Andean
region of the Amazon basin. Paper read at workshop on
Changes to Hydrologic Extremes. IAHS, Perugia (Italy, July
2007).
Espinoza, J.C. et al., 2009b. Spatio–temporal rainfall variability in the
Amazon Basin countries (Brazil, Peru, Bolivia, Colombia and
Ecuador. International Journal of Climatology, 29, 1574–1594.
Farr, T.G. et al., 2007. The Shuttle radar topography mission. Review
of Geophysics, 45, RG2004.
Figueroa, S.N. and Nobre, C., 1990. Precipitation distribution over
central and western tropical South America. Climanalise, 5,
36–44.
Foley, J.A. et al., 2002. El Niño-Southern Oscillation and the climate,
ecosystems and rivers of Amazonia. Global Biogeochemical
Cycles, 16 (4), 1132–1152.
Francou, B. and Pizarro, L., 1986. El Niño et la sécheresse dans les
hautes Andes centrales: étude d’une possible corrélation. Les
Cahiers d’Outre Mer, 156, 371–390.
Garreaud, R.D., 1999. Multiscale analysis of the summertime pre-
cipitation over the Central Andes. Monthly Weather Review,
127 (5), 901–921.
Garreaud, R.D. and Aceituno, P., 2001. Interannual rainfall variabil-
ity over the South American Altiplano. Journal of Climate, 14,
2779–2789.
Garreaud, R., Rutllant, J. and Fuenzalida, H., 2002. Coastal lows
in north-central Chile: mean structure and evolution. Monthly
Weather Review, 130, 75–88.
Garreaud, R., Vuille, M. and Clement, A.C., 2003. The climate of the
Altiplano: observed current conditions and mechanisms of past
changes. Palaeogeography, Palaeoclimatology, Palaeoecology,
194, 5–22.
Garreaud, R.D. et al., 2009. Present-day South American cli-
mate. Palaeogeography, Palaeoclimatology, Palaeoecology,
281 (3–4), 180–195.
Gentry, A.H. and Lopez-Parodi, J., 1980. Deforestation and increased
flooding of the Upper Amazon. Science, 210, 1354–1356.
Guyot, J.-L., 1993. Hydrogéochimie des fleuves de l’Amazonie bolivi-
enne. Paris: Editions de l’ORSTOM.
Herreros, J. et al., 2009. Environmental records from temper-
ate glacier ice on Nevado Coropuna saddle, southern Peru.
Advances in Geoscience, 22, 27–34.
Hiez, G., 1977. L’homogénéité des données pluviométriques. Cahiers
ORSTOM, série Hydrologie, 14, 129–172.
Johnson, A., 1976. The climate of Peru, Bolivia and Ecuador. In:
Climates of Central and South America. World survey of cli-
matology 12, 147–218. New York: Elsevier.
Kane, R.P., 1999. Rainfall extremes in some selected parts of
Central and South America: ENSO and other relation-
ships reexamined. International Journal of Climatology, 19,
423–455.
Kaser, G. et al., 2003. The impact of glaciers on the runoff and the
reconstruction of mass balance history from hydrological data
in the tropical Cordillera Blanca, Peru. Journal of Hydrology,
282, 130–144.
Kendall, M., 1975. Rank correlation methods. London: Charles
Griffin.
Kundzewicz, Z.W. and Robson, A.J., 2004. Change detection in
hydrological records: a review of the methodology: detecting
change in hydrological data. Hydrological Sciences Journal,
49 (1), 7–19.
Laraque, A. et al., 2007. Hydropluviometric regime singularities and
the Ecuadorian Amazon basin. Journal of Hydrometeorology,
8, 1364–1381.
Lavado, W., 2010. Modélisation du bilan hydrique à pas de temps
mensuel pour l’évaluation de l’impact du changement cli-
matique dans le bassin Amazonien du Pérou. Thesis (PhD),
Université Toulouse 3 Paul Sabatier, Toulouse France.
Mann, H.B., 1945. Non-parametric test against trend. Econometrica,
13, 245–259.
Marengo, J.A., 1995. Variations and change in South American
streamflow. Climatic Change, 31, 99–117.
Marengo, J.A., Tomasella, J. and Uvo, C.R., 1998. Trends in
streamflow and rainfall in tropical South America: Amazonia,
eastern Brazil, and northwestern Peru. Journal of Geophysical
Research, 103, 1775–1784.
642 Waldo Sven Lavado Casimiro et al.
Marengo, J.A., 2004. Interdecadal variability and trends of rainfall
across the Amazon basin. Theoretical and Applied Climatology,
78, 79–96.
Mark, B.G., Seltzer, G.O., 2003. Tropical glacier meltwater contribu-
tion to stream discharge: a case study in the Cordillera Blanca,
Peru. Journal of Glaciology, 49, 165, 271–281.
Mark, B.G., McKenzie, J. and Gomez J., 2005. Hydrochemical eval-
uation of changing glacier meltwater contribution to stream
discharge: Callejon de Huaylas, Peru. Hydrological Sciences
Journal, 50 (6), 975–987.
Mark, B.G. et al., 2010. Climate change and tropical Andean
glacier recession: evaluation hydrological changes and liveli-
hood vulnerability in the Cordillera Blanca, Peru. Annals of the
Association of American Geographers, 100, 794–805.
Melice, J. and Roucou, P., 1998. Decadal time scale variability
recorded in the Quelccaya summit ice core 18O isotopic ratio
series and its relation with the sea surface temperature. Climate
Dynamics, 14, 117–132.
Nordin, C.F. et al., 1982. Deforestation and increased flooding of the
Upper Amazon. Science, 215, 426–427.
Oliver, M.A. and Webster, R., 1990. Kriging: a method of interpola-
tion for geographical information system. International Journal
of Geographical Information Systems, 4, 313–332.
Pettitt, A.N., 1979. A non-parametric approach to the change-point
problem. Applied Statistics, 28, 126–135.
Pouyaud, B. et al., 2005. On the future of the water resources from
glacier melting in the Cordillera Blanca, Peru. Hydrological
Sciences Journal, 50 (6), 999–1022.
Poveda, G., 2004. La hidroclimatología de Colombia: una síntesis
desde la escala inter-decadal hasta la escala diurna. Revista de
la Academia Colombiana de Ciencias, 28, 201–222.
Pulwarty, R.S., Barry, R.G. and Riehl, H., 1992. Annual and seasonal
patterns of rainfall variability over Venezuela. Erdkunde, 46,
273–289.
Richey, J.E., Nobre, C. and Deser, C., 1989. Amazon River dis-
charge and climate variability: 1903 to 1985. Science, 246,
101–103.
Rigsby, C.A., Baker, P.A. and Aldenderfer, M.S., 2003. Fluvial
history of the Rio Ilave valley, Peru, and its relationship to cli-
mate and human history. Palaeogeography, Palaeoclimatology,
Palaeoecology, 3061, 1–21.
Robson, A. et al., 2000. Statistical methods for testing for change.
Chapter 5 in: Z.W. Kundzewicz and A. Robson, eds., Detecting
trend and other changes in hydrological data. Geneva: World
Meteorological Organization, WMO/TD 1013.
Rocha, H.R.d., Nobre, C.A. and Barros, M.C., 1989. Variabilidade
natural de longo prazo no ciclo hidrologico da Amazonia.
Climanálise, 4, 36–42.
Roche, M. et al., 1990. Hétérogénéité des précipitations sur la cordil-
lère des Andes boliviennes. In: H. Land and A. Musy, eds.,
Hydrology in mountainous regions I – Hydrological measure-
ments, the water cycle. Wallingford, UK: IAHS Press, IAHS
Publ. 193, 381–388.
Rome-Gaspaldy, S. and Ronchail, J., 1998. La pluviométrie au
Pérou pendant les phases ENSO et LNSO. Bulletin de l’Institut
Française d’Etudes Andines, 27, 675–685.
Romero, C.C., Baigorria, G.A. and Stroosnijder, L., 2007. Changes
of erosive rainfall for El Niño and La Niña years in the
northern Andean highlands of Peru. Climatic Change, 85,
343–356.
Ronchail, J. et al., 2002. Interannual rainfall variability in the Amazon
basin and sea-surface temperatures in the equatorial Pacific
and the tropical Atlantic Oceans. International Journal of
Climatology, 22, 1663–1686.
Ronchail, J. and Gallaire, R., 2006. Enso and rainfall along the
Zongo valley (Bolivia) from the Altiplano to the Amazon basin.
International Journal of Climatology, 26, 1223–1236.
Ropelewski, C.F. and Halpert, M.S., 1987. Global and regional
scale precipitations patterns associated with El Nino-Southern
Oscillation. Monthly Weather Review, 115, 1606–1626.
Ruiz, R.R., Torres, H.G. and Aguirre, M.N., 2008. Delimitación y
codificación de unidades hidrográficas del Perú. Lima, Peru:
Autoridad Nacional del Agua-Ministerio de Agricultura.
Rutllant, J., Fuenzalida, H. and Aceituno, P., 2003. Climate dynam-
ics along the arid northern coast of Chile: the 1997–1998.
Dinámica del clima de la Región de Antofagasta (DICLIMA)
experiment. Journal of Geophysical Research, 108 (D17),
4538–4552.
Sicart, J.E., 2002. Contribution à l’étude des flux d’énergie, du bilan
de masse et du débit de fonte d’un glacier tropical: Le Zongo,
Bolivie. Thèse de doctorat, Univ. Paris 6, France.
Tapley, T.D. and Waylen, P.R., 1989. A mixture model of annual
precipitation in Peru. The Professional Geographer, 41,
62–71.
Tapley, T.D.J. and Waylen, P.R., 1990. Spatial variability of annual
precipitation and ENSO events in western Peru. Hydrological
Sciences Journal, 35, 429–445.
UNESCO, 2006. Balance hídrico superficial del Perú a nivel multian-
ual. Montevideo: Documentos Técnicos del PHI-LAC, no. 1.
Vauchel, P., 2005. Hydraccess: Logiciel de gestion et traitement de
données hydro météorologiques, version 2.1.4. Downloadable
at http://www.mpl.ird.fr/hybam.01/01/2011
Vizy, E.K. and Cook, K.H., 2007. Relationship between Amazon
and high Andes rainfall. Journal of Geophysical Research—
Atmospheres, 112 (D7) 14 pp. doi:10.1029/2006JD007980
Vuille, M. et al., 1998. Atmospheric circulation anomalies associated
with 1996/1997 summer precipitation events on Sajama ice
cap, Bolivia. Journal of Geophysical Research—Atmospheres,
103 (D10), 11191–11204.
Vuille, M., Bradley, R.S. and Keimig, F., 2000. Interannual climate
variability in the Central Andes and its relation to tropical
Pacific and Atlantic forcing. Journal of Geophysical Research—
Atmospheres 105, 12447–12460.
Vuille, M. et al., 2008a. Climate change and tropical Andean
glaciers: Past, present and future. Earth-Science Reviews, 89,
79–96.
Vuille, M., Kaser, G. and Juen, I., 2008b. Glacier mass balance vari-
ability in the Cordillera Blanca, Peru and its relationship with
climate and the large-scale circulation. Global and Planetary
Change, 62, 14–28.
Wagnon, P. et al., 1999. Annual cycle of energy balance of
Zongo glacier, Cordillera Real, Bolivia. Journal of Geophysical
Research, 104 (D4), 3907–3923.
Wagnon, P. et al., 2001. Anomalous heat and mass budget of Glaciar
Zongo, Bolivia, during the 1997–98 El Niño year. Journal of
Glaciology, 47 (156), 21–28.
Waylen, P. and Poveda, G., 2002. El Niño-Southern Oscillation
and aspects of western South American hydro-climatology.
Hydrological Processes, 16 (6), 1247–1260.
Waylen, P.R. and Caviedes, C.N., 1986. El Niño and annual floods
on the north Peruvian littoral. Journal of Hydrology, 89 (1–2),
141–156.
Yoon, J.H. and Zeng, N., 2009. An Atlantic influence on Amazon
rainfall. Climate Dynamics, 34 (2–3), 249–264.
Zhang, X.B. et al., 2001. Trends in Canadian streamflow. Water
Resources Research, 37 (4), 987–998.
Zhou, J. and Lau, K-M., 1998. Does a monsoon climate exist over
South America? Journal of Climate, 11, 1020–1040.