Fortyeight

Quantifying the significance of recent glacier recession in the Cordillera Blanca, Perú: a case study of hydrological impact and climatic control

Bryan G. Mark1

Department of Geography, The Ohio State University, Columbus, OH 43210, USA

48.1 Introduction

The Peruvian Cordillera Blanca is the most extensively glacierized mountain range in the tropics, and an important location to study the practical impact and climatic control of ongoing glacier volume loss. Draining most of the glacierized area in the Cordillera Blanca, the Río Santa (Fig. 48.1a) maintains the second largest and least variable annual discharge of all rivers in Perú flowing to the Pacific Ocean, from which the economically developing population derives water. If the glaciers disappear, as has been predicted for other tropical regions, important climatic archives will be lost, and the region could face a future water supply crisis. A programme of routine surveying was established in 1968, and four target glaciers were selected to monitor the magnitude of changes to these frozen reservoirs, providing a particularly important geographical database (Ames, 1998). Moreover, gauges in tributary basins discharging to the upper reaches of the Río Santa, known as the Callejon de Huaylas (Fig. 48.1a), provide

'Formerly at Max Planck Institute for Biogeochemistry, Jena 07745 Germany and Department of Geography & Geomatics, University of Glasgow, Glasgow G12 8QQ, UK

a 50-yr record of runoff and precipitation. This case study focuses first on one of these target glaciers, Glaciar Yanamarey, where the magnitude of recession is known, then quantifies the downstream impact of meltwater.

48.2 Quantifying 50 yr of net glacier recession: Glaciar Yanamarey

Data from GPS surveys of Glaciar Yanamarey extend the record of mapped terminus positions to over three decades at this site, where early quantitative evaluations of modern tropical glacier volume loss were made (Hastenrath & Ames, 1995a,b) (Fig. 48.1c). Although the terminus experienced some small advances, the late 20th century featured extensive glacier recession, consistent with other tropical glaciers and pervasive wastage of glaciers globally. Recession rates also increased, as in other Andean locations. These length variations express a rapid response to massbalance changes, captured by the basin hydrological balance (Kaser et al., 2003). Using a simplified annual mean budget based on a decadal-scale velocity and net-balance data, Hastenrath & Ames (1995b) estimated that about 50% of the water discharging from Glaciar Yanamarey was not renewed by precipitation, but provided by progressive thinning, and that the receding glacier

Figure 48.1 Case study location maps of successively larger scale: (a) Callejón de Huaylas, a watershed of ca. 5000 km2 draining the Cordillera Blanca, Perú, to the upper Río Santa. Stream-gauge locations mentioned in the text are identified; (b) Que-rococha watershed, 60 km2, showing the discharge and water sampling points mentioned in the text; (c) Yanamarey catchment, 1.3 km2 between 4600 m and 5300 m, 75% of which is covered by glacier ice. The shaded region shows the outline of Glaciar Yanamarey in 1982, with contours and a centre-line to show distance from the headwall at 100-m intervals (after Hastenrath & Ames, 1995a). Terminus positions are mapped onto a common datum, based on surveys for 1939,1948,1962,1973, 1982,1988, 1997, 1998 and 1999. The latter three positions were mapped using differential GPS. The cumulative terminus recession from the 1939 position is shown (m) on the inset graph as a solid line, with solid rectangles for years with corresponding terminus position mapped (data from A. Ames, personal communication, 1998), along with average recession rate between years with mapped termini (in myr-1). Asterix marks the location of a weather station, where daily temperature and monthly precipitation were recorded discontinuously from 1982.

Figure 48.1 Case study location maps of successively larger scale: (a) Callejón de Huaylas, a watershed of ca. 5000 km2 draining the Cordillera Blanca, Perú, to the upper Río Santa. Stream-gauge locations mentioned in the text are identified; (b) Que-rococha watershed, 60 km2, showing the discharge and water sampling points mentioned in the text; (c) Yanamarey catchment, 1.3 km2 between 4600 m and 5300 m, 75% of which is covered by glacier ice. The shaded region shows the outline of Glaciar Yanamarey in 1982, with contours and a centre-line to show distance from the headwall at 100-m intervals (after Hastenrath & Ames, 1995a). Terminus positions are mapped onto a common datum, based on surveys for 1939,1948,1962,1973, 1982,1988, 1997, 1998 and 1999. The latter three positions were mapped using differential GPS. The cumulative terminus recession from the 1939 position is shown (m) on the inset graph as a solid line, with solid rectangles for years with corresponding terminus position mapped (data from A. Ames, personal communication, 1998), along with average recession rate between years with mapped termini (in myr-1). Asterix marks the location of a weather station, where daily temperature and monthly precipitation were recorded discontinuously from 1982.

would survive another 50 yr in the present climate. Thus, a quantitative estimate of how much the downstream hydrology is impacted by glacier wastage is required for water resource planning.

48.3 Quantifying downstream impact of glacier meltwater: Callejon de Huaylas

Monthly observations of specific precipitation and discharge (P and Qt, respectively) were collected with hydrochemical samples over the 1998-1999 hydrological year at the Yanamarey glacier catchment (YAN) and in the larger Querococha watershed downstream, where a confluence of glacierized (YAN, Q2) and non-glacierized (Q1) streams forms a tributary stream to the Río Santa (Q3) (Fig. 48.1b). Maximum Qtprecedes the peak in average Pby 4 months at YAN, whereas Qt is diminished and closely correlated in time to P at Q3 (Fig. 48.2). A simple water-balance calculation shows that the maximum in specific melt occurs in October for Glaciar Yanamarey, and the April minimum is negative, representing net accumulation (Fig. 48.2a). However, melt contributes a maximum relative percentage of the monthly Qt during the dry season months (June-September). During this period of little to no precipitation, glacier melt contributes up to 100% of Qt, thereby buffering the downstream flow. Assuming that the loss in glacier storage is exclusively by melting, then glacier meltwater comprises 35-45% of the total annual stream discharge from YAN. The Querococha watershed is a good analogue for the entire Callejon de Huaylas, as both are about <10% glacierized. Here, the relative influence of glacier meltwater to the annual runoff regime diminishes downstream of YAN at Q2, to become precipitation dominated at Q3, as in the non-glacierized Q1 (Fig. 48.2b). A volume-weighted hydrochemical mixing model (using dissolved anions and cations) revealed that YAN contributes about one-third of the discharge at Q3 over an annual cycle (Mark & Seltzer, 2003). By analogy, the larger Río Santa watershed thus

700 600 500 400 300

J 200 100 0

300 P (b) Querococha 200

00

00

00

ON

ON

On

n

ul

CD

u

u

0

00

00

00

00

ON

ON

ON

ON

ON

ON

ON

ON

ON

ON

ON

ON

ON

On

o

>

c

r

r

n

n

CO

o z

e o

Ll_

a

CP <

u

u

ON

ON

2

2

0

5 0

3 2

2 0

(c) Callejón de Huaylas

L Querococha '9

Olleros

20 40

Glacierized area (%)

Figure 48.2 Hydrological and climatological data from the successively larger catchments of the case study (see Fig. 48.1): (a) observational data from the Yanamarey glacier catchment, including monthly measurements of specific discharge (Qt) (mm) from YAN plotted with the monthly precipitation totals (P) (mm) and monthly average temperature (T) (°C) sampled over the 1998-1999 hydrological year, plotted with the glacier melt (Melt) calculated from a simplified hydrological mass balance; (b) specific discharge data from locations in the Querococha watershed plotted with monthly precipitation at the Querococha gauge (both in mm), on the same scale as (a); (c) magnitude and variation of annual stream discharge with percentage of glacierized area in the Río Santa tributaries, shown by ratio of maximum monthly discharge to mean monthly discharge (max Q / mean Q); labelled data points correspond to gauge locations shown in Fig. 48.1a.

-200

-400

-600

-200

-400

-600

3.0 2.0 1.0 0.0 £ -1.0 -2.0 -3.0

2

3

4

6

K

00

ON

0

1

2

3

4

ON

00

00

00

00

00

00

00

00

ON

ON

ON

ON

ON

ON

1-

2-

3

-4

-6

hL

00

On

-0

1-

-2

3

00

00

00

00

00

00

00

00

00

00

ON

ON

ON

ON

ON

ON

ON

ON

ON

ON

ON

ON

ON

ON

ON

ON

ON

ON

ON

1

1

1

1

1

1

1

1

1

1

1

1

1

1

Figure 48.3 Total annual precipitation (P) (in mm) and mean annual temperature (in °C) anomalies for the Yanamarey catchment during years with complete data (1981-1994, 1998-1999). Monthly measurements were aggregated over the course of a hydrological year (June-May). Temperature data are shown with one standard deviation error bars based on monthly means from the weather station in Yanamarey catchment (YT) where available, and corrected using lapse rate computed from 1956 to 1997 time series measured at the Querococha stream gauge (QT). Observations are incomplete after 1994, and only reestablished in 1998, as denoted by the break in the x axis.

receives a significant amount (10-20%) of its annual discharge from melting glacier ice. Historical records of discharge from tributary basins of the Callejon de Huaylas confirm that watersheds feature enhanced mean annual discharge and less variable runoff in proportion to glacierized area (Fig. 48.2c).

48.4 Insights into climatic control

The tropical austral spring climate during maximum glacier discharge is conducive to melting ice, emphasizing the hydrological and climatological significance of these transitional months between the late dry and early wet seasons. Mass balance and basin hydrology are tied most closely to atmospheric conditions during these spring-summer months, when a lower surface albedo absorbs more long-wave energy and an increase in humidity shifts the latent energy balance at the glacier surface towards producing more melting than sublimation/evaporation. Thus cloudiness, humidity and precipitation probably have more control on melt in the short term (Francou et al., 2003). This confirms a combination of climatic variables originally hypothesized to have forced 20th century recession at Glaciar Yanamarey (Hastenrath & Ames, 1995b). Recent research on tropical glaciers and runoff in Bolivia and Ecuador (e.g. Francou et al., 2000) has shown that total net all-wave radiation, not temperature, is the main factor controlling ablation. However, because temperature is strongly interconnected with these variables, it remains an important indicator of longer term glacier evolution linked to large-scale climate processes such as ENSO (e.g. Vuille et al., 2003; Kaser et al., 2003). Observational evidence from Andean meteorological stations shows a strong positive trend in temperature (Mark, 2002), and model results confirm that the century-scale Andean glacier recession is best explained by increased temperature and humidity (Vuille et al., 2003). Temperature measurements from the Yanamarey catchment show that discharge is actually better correlated with temperature than precipitation (Fig. 48.2a). Moreover, discontinuous time series show an intriguing association of higher temperature with increased precipitation during at least the 1982, 1994 and 1997-1998 El Niño events (Fig. 48.3), opposite from the expected deficit of precipitation driving enhanced ablation during El Niño further south along the Andes (e.g. Wagnon et al., 2003). Downstream, the seasonal contrast in discharge due to a strong amplitude precipitation signal is mitigated by glacier melt, distinct from mid-latitude glaciers responding to strong seasonal temperature contrasts (Mark & Seltzer, 2003).

0 0

Post a comment