Hydrological response to climate change in a glacierized catchment in the Himalayas
The analysis of climate change impact on the hydrology of high altitude glacierized catchments in the Himalayas is complex due to the high variability in climate, lack of data, large uncertainties in climate change projection and uncertainty about the response of glaciers. Therefore a high resolution combined cryospheric hydrological model was developed and calibrated that explicitly simulates glacier evolution and all major hydrological processes. The model was used to assess the future development of the glaciers and the runoff using an ensemble of downscaled climate model data in the Langtang catchment in Nepal. The analysis shows that both temperature and precipitation are projected to increase which results in a steady decline of the glacier area. The river flow is projected to increase significantly due to the increased precipitation and ice melt and the transition towards a rain river. Rain runoff and base flow will increase at the expense of glacier runoff. However, as the melt water peak coincides with the monsoon peak, no shifts in the hydrograph are expected.
More than one-sixth of the global population rely on glacier and snow melt for their water supply (Barnett et al. 2005). Changes in temperature and precipitation are expected to significantly affect the cryospheric processes and the hydrology of headwater catchments in the Himalayas (Cruz et al. 2007; Immerzeel et al. 2009). Accurate modeling of the hydrological response to climate change in these catchments is complicated due to the large climatic heterogeneity, the lack of data in mountain catchments, the resolution and accuracy of GCM outputs and uncertainty about the response of glacier and snow dynamics (Beniston 2003). Traditionally, glacier dynamics are not linked to other hydrological processes such as evapotranspiration, surface runoff and base flow in a single model (Sharp et al. 1998). Most hydrological impact studies on the contrary deploy simple degree day methods (Hock 2005) and assume hypothetical reduction in future glacier areas (Singh and Bengtsson 2004; Hock 2005; Rees and Collins 2006; Immerzeel et al. 2009). Although these studies provide valuable insights into the possible range of future options, they suffer from large uncertainty about the plausibility of the future evolution of snow and ice. Another issue that is often ignored in modeling melt from glaciers is that both accumulation and ablation of glaciers depend on the glacier area, which does not scale linearly with glacier volume. As larger glaciers have a smaller area to volume ratio they are less sensitive to climate change than smaller glaciers (Van de Wal and Wild 2001).
Recently there has been a strong debate about the melt rate of Himalayan glaciers. The claim that all glaciers in the Himalayas could disappear by 2035 in the AR4 of the IPCC (Cruz et al. 2007) has been the source of great controversy and has been admitted to stem from a wrong quotation of grey literature (Schiermeier 2010). Mass balance studies and regional reviews conclude otherwise, although there are very limited published studies and measurements available (Bolch et al. 2008; Zemp et al. 2009). Obviously, there is large uncertainty and variability in the retreat rates of Himalayan glaciers and to settle the debate in a rational manner there is a strong need for reference studies that reliably model the transient evolution of glaciers under climate change.
In this study we attempt to provide such a reference study by developing a combined cryospheric hydrological model for a glacierized Himalayan catchment in Nepal. The model explicitly simulates glacier movement in combination with major hydrological processes such as evapotranspiration, surface runoff, ablation and groundwater base flow at a high spatial resolution with a daily time step. An innovative two stage calibration procedure is used. First the historical evolution of glaciers is calibrated using the recent location of glacier tongues and secondly the hydrological processes are optimized using discharge observations. The calibrated model is then forced with an ensemble of transient statically downscaled GCM outputs to evaluate the climate change impact on the future evolution of the glaciers and catchment hydrology.
2 Study area
The entire model is developed using the PCRaster environment for numerical modeling in environment science (Karssenberg et al. 2001). The model is setup at a spatial resolution of 90 m (338 × 325 cells). For each cell the following model concepts are simulated at a daily time step.
Equilibrium shear stress
Material roughness coefficient
N m-2 s1/3
Threshold temperature for precipitation to fall as snow
Temperature lapse rate
Degree day factor
mm °C day-1
Aspect dependence of ddf
Multiplicative factor for the ddf for debris covered glaciers
Multiplicative factor for the ddf for the Lirung glacier (exceptional thick debris cover)
Runoff fraction of total ablation
Vertical precipitation lapse rate
Horizontal precipitation lapse rate (negative from west to east)
Temperature correction factor reference evapotranspiration
Reduction factor to calculate Eta
Maximum soil water capacity
3.2 Hydrological processes
The model is forced by daily precipitation and temperature data at Kyangjing from 2000 to 2007. Temperature is spatially differentiated using a vertical lapse (λt) rate, which is a calibration parameter. For precipitation both a positive vertical (λp,v) and a negative horizontal gradient (λp,h) from west to east is applied similar to (Konz et al. 2007). Precipitation is partitioned in either snow (S) or rain (P) using these daily fields, the lapsed temperature fields and a threshold temperature.
Reference evapotranspiration (ET0) is calculated based on minimum, maximum and average temperature data at Kyangjing from 2000 to 2007 according to the Hargreaves equation (Hargreaves et al. 1985). ET0 is spatially differentiated per cell by applying a temperature dependent correction factor (cET0). This factor is derived by using the temperature dependence of the Hargreaves equation. Actual ET (ETa) is derived by first deriving potential evapotranspiration by multiplying ET0 by a crop reduction factor (Ke). By limiting potential evapotranspiration with the actual soil water content actual evapotranspiration is calculated.
Where k is a recession coefficient that is calibrated and Qout,t (m3 s-1) is the river discharge at the catchment outlet on day t. The model parameters related to the hydrological processes are shown in Table 1.
The model is calibrated using the Parameter ESTimation (PEST) software. PEST is able to run a model as many times as it needs to while adjusting its parameters until the discrepancies between selected model outputs and a set of observations is reduced to a minimum. The PEST algorithm method is based on non-linear parameter estimation theory (Doherty 2005).
The calibration procedure follows a two step approach. First the parameters related to the glacier modeling are manually calibrated as glacier evolution is a much slower process than the hydrological processes such as evapotranspiration, surface runoff, ablation and base flow. Four parameters that influence glacier evolution are calibrated (τ0, R, λt, ddf). A bias corrected time series of precipitation and temperature from 1957–2002, based on the ERA40 dataset (Uppala et al. 2005), is used to force the model with the aim to reproduce the location of glaciers and permanent snow in the year 2000 based on remote sensing (Konz et al. 2007). The bias correction is done similar to the corrections applied for the GCMs Eq. 7/Eq. 9. In the second step, PEST is used and four hydrological parameters are calibrated (θm, CN, α and k) by simulating the period 2000–2006 for which observations of precipitation and temperature at Kyangjing are available. The parameters are calibrated against observed daily discharges from 2000–2006 at the outlet of the catchment.
3.4 Downscaling GCM output
Where a’gcm, M is the corrected climate parameter.
The monthly downscaled GCM time series from 2001 to 2099 were subsequently disaggregated into daily values by using monthly distributions of the daily 45 year ERA40 dataset from 1957–2002. The total monthly sums are equal to the downscaled GCM time series but the partitioning into daily values is based on the statistical distribution of the ERA40 dataset. This procedure resulted in a daily statistically downscaled precipitation and temperature time series from 2001 to 2099 for the five different GCMs for the Kyangjing station, which were then used to force the calibrated model. It can be expected that in the transient GCM runs both the effects as climate change (trends) as well as multi-year variability are present, such that they will be accounted for in our projections. Finally using ERA40 to simulate daily variation will assure that within month variation is properly represented. Note that further distribution of precipitation and temperature of the downscaled Kyangjing time series within the catchment is achieved by applying vertical temperature lapse rate and horizontal and vertical precipitation lapse rates (Table 1) similar to the reference run from 2000–2007.
4 Results and discussion
The combination of hydrological processes and glacier movement in a high resolution raster based model enables the accurate simulation of both glacier evolution and river flow in a high altitude glacierized catchment in the Himalayas.
Climate change analysis using downscaled data from 5 different GCMS shows that temperatures are projected to increase by 0.06°C y-1 and precipitation by 1.9 mm y-1. The analysis also reveals a large variability among the different GCMs in particular for precipitation.
In the catchment the glaciers are retreating steadily under climate change and it is estimated that in 2035 the glacier area has been reduced by 32%. This catchment is representative for the southern slopes of central and eastern Himalayas where glacier systems are dynamic, moderate in size and often characterized by debris covered tongues.
The positive temperature and precipitation trends will increase evapotranspiration and snow and ice melt while more precipitation will fall as rain instead of snow. The net result is an increase in stream flow by 4 mm y-1 that can be attributed to the increase in precipitation and the change from melt river to rain river. The partitioning of stream flow is indeed showing strong changes. Rain runoff and base flow are increasing, snow runoff remains more or less constant and glacier runoff is eventually decreasing.
This study shows an extensive analysis of a glacierized catchment in the central Himalayas but the results are not representative for the entire Himalayas. To arrive at a comprehensive assessment on how climate change is affecting the hydrology of the Himalayas it is recommendable to perform this analysis in reference catchments covering the east–west and north–south gradient in climatology and glacier and hydrological dynamics.
This study was financially supported by the Netherlands Organisation for Scientific Research (NWO) through a CASIMIR grant (018 003 002) and by the European Commission (Call FP7-ENV-2007-1 Grant nr. 212921) as part of the CEOP- AEGIS project (http://www.ceop-aegis.org/) coordinated by the Université de Strasbourg.
This article is distributed under the terms of the Creative Commons Attribution Noncommercial License which permits any noncommercial use, distribution, and reproduction in any medium, provided the original author(s) and source are credited.
- Ageta Y, Higuchi K (1984) Estimation of mass balance components of a summer-accumulation type glacier in the Nepal Himalaya. Geografiska Annaler. Series A, Physical Geography, 66(3), p.249. Available at: http://www.jstor.org/stable/520698?origin=crossref.
- Cruz R et al (2007) Asia. In: Parry ML et al (eds) Climate change 2007: impacts, adaptation and vulnerability. contribution of working group II to the fourth assessment report of the intergovernmental panel on climate change. Cambridge University Press, CambridgeGoogle Scholar
- Doherty J (2005) PEST: model independent parameter estimation 5th ed. Watermark Numerical Computing, BrisbaneGoogle Scholar
- Hargreaves GL, Hargreaves GH, Riley JP (1985) Agricultural benefits for Senegal River Basin. Irrigation and Drainage Engr., ASCE 111, pp 113–124Google Scholar
- Kumar S, Dobhal DP (1997) Climatic effects and bedrock c ontrol on rapid fluctuations of Chhota Shigri glacier, northwest Himalaya, India. J Glaciol 43:467–472Google Scholar
- Mae S et al (1975) Thermail drilling and temperature measurements in Khumbu Glacier, Nepal Himalayas. Seppyo 37(161–169)Google Scholar
- Martinec J (1975) Snowmelt-runoff model for stream flow forecasts. Nord Hydrol 6:145–154Google Scholar
- Meehl GA et al (2007) Global climate projections. In: Miller HL, Solomon S, Tignor M et al (eds) Climate change 2007: the physical science basis. contribution of working group i to the fourth assessment report of the intergovernmental panel on climate change. Cambridge University Press, CambridgeGoogle Scholar
- Neitsch SL et al (2005) Soil and water assessment tool: theoretical documentation. Agricultural Research Service, TempleGoogle Scholar
- Randall DA et al (2007) Climate models and their evaluation. In: Miller HL, Solomon S, Tignor M et al (eds) Climate change 2007: the physical science basis. contribution of working group I to the fourth assessment report of the intergovernmental panel on climate change. Cambridge University Press, CambridgeGoogle Scholar
- Sakai A, Fujita K, Kubota J (2004) Evaporation and percolation effect on melting at debris-covered Lirung Glacier, Nepal Himalayas, 1996. Bull Glacier Res 21:9–15Google Scholar
- Schiermeier Q (2010) Glacier estimate on thin ice. Geophys Res Lett 463(21):276–277Google Scholar
- SCS USDA (1972) National engineering handbook section for hydrology. In Washington, D.C.Google Scholar
- Seko K (1987) Seasonal variation of altitudinal dependence of precipitation in Langtang Valley, Nepal Himalayas. Bull Glacier Res 5:41–47Google Scholar
- Sharp M, Richards KS, Tranter M (1998) Advances in hydrological processes; glacier hydrology and hydrochemistry. Wiley, ChichesterGoogle Scholar
- Shiraiwa T, Ueno K, Yamada T (1992) Distribution of mass input on glaciers in the Langtang Valley, Nepal Himalayas. Bull Glacier Res 10:21–30Google Scholar
- Smedema LK, Rycroft DW (1983) Land drainage—planning and design of agricultural drainage systems. Cornell University Press, IthicaGoogle Scholar
- Van de Wal RSW, Wild M (2001) Modelling the response of glaciers to climate change by applying volume-area scaling in combination with a high resolution GCM. Clim Dynam 18(3–4):359–366Google Scholar
- Wang B (2006) The Asian monsoon. Springer, BerlinGoogle Scholar
- Weertman J (1957) On the sliding of glaciers. J Glaciol 3:33–38Google Scholar
- Zemp M, Hoelzle M, Haeberli W (2009) Six decades of glacier mass-balance observations: a review of the worldwide monitoring network. Annals of Glaciology, 50(50), pp 101–111. Available at: http://openurl.ingenta.com/content/xref?genre=article&issn=0260-3055&volume=50&issue=50&spage=101