Elsevier

Journal of Hydrology

Volume 522, March 2015, Pages 239-249
Journal of Hydrology

Evaporative fractions and elevation effects on stable isotopes of high elevation lakes and streams in arid western Himalaya

https://doi.org/10.1016/j.jhydrol.2014.12.023Get rights and content

Highlights

  • Stable isotopes of oxygen and hydrogen constrain water balance of Himalayan lakes.

  • Elevation–isotope relationship in streams changed little with elevation.

  • Largest lakes showed high evaporation and little loss to outflow.

  • Little isotopic evidence of hydrologic contribution of enriched lakes to Indus River.

Summary

Isotopes of oxygen and hydrogen in water from streams, snow, and lakes were used to model the ratio of evaporation to total inflow (E/I) of four high elevation lakes in closed basins in the Indian Himalaya. Air temperature and relative humidity (h) data from meteorological stations and global climate grids (GMAO-MERRA) were used as input to the model. A second model of the volume of inflow during snowmelt constrained the magnitude of seasonal variability in isotopic composition. Similar to other areas of the Himalaya, elevation was a strong determinant of isotopic composition of stream water, suggesting that heavier isotopes rain out at lower elevations. Deuterium excess (d) in stream water suggests that summer precipitation originating from the Bay of Bengal rather than winter precipitation from Central Asia is the dominant source of precipitation. For the largest and deepest lakes (>15 m), E/I was 77–87%, and 18–50% for two shallow lakes (<2 m). Modelled E/I was sensitive to h; ground-level measurements of h are needed to constrain E/I estimates in arid mountain regions, though h from GMAO also produced reasonable E/I values. The results suggest that some lakes in closed basins in the Himalaya lose a significant fraction of their inflow to groundwater, particularly shallow lakes with a low ratio of lake volume to watershed area. Isotopic values from the mainstem of the Indus River suggest that evaporatively enriched waters have limited impact on both wet and dry season discharge at the basin scale. Lakes with a high evaporative fraction may be uniquely sensitive to climate, and isotopic analysis can help identify lakes that may be vulnerable to climate fluctuations and change.

Introduction

Lakes play important roles in the hydrology and ecology of mountainous regions. Lakes can either increase dry season streamflow by recharging local or regional groundwater, or decrease dry season streamflow by increasing evaporation (Bullock and Acreman, 2003). Significant uncertainty remains about the influence of high elevation lakes on monthly and annual streamflow. The hydrology of lakes in closed basins, which have no surface outlet, can be particularly difficult to determine since the main outflow is through recharge to groundwater. Lakes in closed basins may impact regional hydrology through subsurface flow, which is difficult to quantify, and is usually calculated as a residual in the water balance. Given the likelihood of significant changes in the hydrology of the Himalaya due to climate change (Immerzeel et al., 2010), it is imperative to understand the role that high elevation lakes play in the seasonal and annual water balance of Himalayan river systems. There is currently very little information on the role of evaporation and subsurface flow in the annual hydrologic budget of Himalayan lakes in closed basins. An interannual water balance on one of the largest lakes in the Tibetan Plateau (Nam Co) suggests that more than half of the inflow to the lake is lost through groundwater seepage, with important implications for the water balance of the Tibetan Plateau and receiving water bodies (Zhou et al., 2013).

Evaporation as a fraction of inflow to a lake (E/I) can be calculated using oxygen (18O) and deuterium (D) isotopes and a model of isotopic fractionation during evaporation (Dincer, 1968, Gibson and Edwards, 2002, Krabbenhoft et al., 1990, Zuber, 1983). E/I models are based on the evaporative enrichment of 18O and D, which evaporate at lower rates than 16O and 1H. The departure of δ18O and δD in a sample from the global meteoric water line (GMWL) is proportional to E/I. Isotopic techniques are particularly useful where there are no discharge measurements to constrain a water balance, as is the case for many remote regions. Applications of E/I models include regional sampling in the Canadian Arctic (Gibson and Edwards, 2002, Gibson et al., 2002), where E/I decreased with increasing latitude. While regional surveys of the isotopic composition of water have been conducted in the Himalaya (Bartarya et al., 1995, Pande et al., 2000), and isotopes have been used to quantify surface–groundwater exchange in lakes at the foothills of the Himalaya (Nachiappan and Kumar, 2002), the meteorological data required to perform an isotopic water balance have not typically been available, nor have isotopic measurements been used to estimate E/I for lakes in the Himalaya.

Both stable isotopes of water (18O and D) can be used to estimate E/I and to test the validity of model assumptions. Zuber (1983) applied the E/I model to several lakes that had complementary water balance data, and documented that the field-observed kinetic enrichment of D during evaporation is sensitive to humidity (h), showing significant departures from values determined in laboratory experiments under conditions of moderate to low h (<70%). This is potentially important for modelling the water balance using isotopic methods in semi-arid and arid regions.

The meteorological data required to implement isotopic E/I models include air temperature and h, which may not be readily available in remote regions. Global gridded data, like that from the Global Meteorological Analysis Office (GMAO) could be used as input, though their cell sizes (e.g. 0.67°) may complicate their use in mountainous environments. E/I values may also be impacted by seasonal inflow of isotopically light water that has not been impacted by evaporation, including snowmelt (Gibson et al., 2002). Implementation of E/I models requires an estimate of uncertainty and evaluation of model sensitivity to meteorological inputs and model assumptions.

Isotopes of water have also been used to reconstruct paleoelevations of mountain ranges based on the relationship between elevation and isotope composition in modern and paleowater samples (Garzione et al., 2000). Regional variations in the elevation–isotope relationship complicate such reconstructions (Hren et al., 2009), and the elevation–isotope relationship may weaken at high elevations (Pande et al., 2000), so additional documentation of the elevation–isotope relationship in sparsely-sampled environments is required, particularly at high elevations in remote regions of the Himalaya. Data on 18O and D, in particular the deuterium excess (d) can also help identify the dominant sources of moisture in precipitation, streams and groundwater (Weyhenmeyer et al., 2002). In the Himalaya, they have been used to quantify the relative contribution of moisture from the Bay of Bengal to the east during the Indian Summer Monsoon and from central Asia to the west and north during the winter (Hren et al., 2009, Karim and Veizer, 2002, Maurya et al., 2011).

The objective of this study is to use a combination of water balance modelling, meteorological measurements and isotopic analyses of stream water, lake water and snow to estimate the roles of evaporation and subsurface discharge in the annual water balance of four high elevation lakes located in closed basins in the Indian Himalaya. A secondary objective is to document the regional elevation–slope relationship and deuterium-excess and compare them with other regional studies. The main research questions are: What fraction of the inflow to high elevation lakes is evaporated, estimated using an E/I model and data on 18O and D? How sensitive are E/I estimates to input meteorological data and potential seasonal variations in isotopic composition? Are differences in the evaporative fractions related to lake depth or watershed characteristics? How does stream water isotopic composition change with elevation? What do deuterium excess values in streamflow suggest about the dominant sources of moisture for precipitation?

Section snippets

Study area

The study area includes four lakes in the Ladakh region of the Indian Himalaya in the State of Jammu and Kashmir, south and east of the city of Leh (Fig. 1, Table 1). Two lakes are relatively deep, and two are shallow (Table 1). The first and largest lake, Tso Moriri, lies at an elevation of 4535 m.a.s.l. Its watershed is part of the Sutlej sub-basin of the Indus River. Based on Landsat TM imagery from 1972–2009, the surface area of Tso Moriri ranges from 139.7 to 145.1 km2 (Lesher, 2011)

Isotope evaporation model

Stable isotopes of oxygen (18O) and hydrogen (D) were used to constrain the water balance of the four studied lakes. The methods follow a combination of Gibson and Edwards (2002) and Gibson et al. (2002), which are based on isotopic mass balance. The main objective is to determine the fraction of inflowing water (I) that is evaporated from the lake surface (E/I), as:EI=δL-δIm(δ-δL)where δL is the measured isotopic composition of the lake, δI is the measured isotopic composition of inflow to

Elevation and deuterium excess

The isotopic composition of stream water varied linearly with the elevation of the sample location (Fig. 3, Table 3), which reflects rain out of the heavier isotope at lower elevations. Linear regression models were used to estimate slope of the isotope–elevation relationship for both the 2011 dataset and for the dataset of Pande et al. (2000) who sampled some of the same streams (Fig. 3, Table 4). The lakes all had similar elevations (4535–4736 m), so the modelled isotopic composition of inflow

Isotopic composition in regional context

The deuterium excess (d) of stream water (+7.5‰) was similar to d in precipitation from the summer monsoon (+8‰) (Karim and Veizer, 2002), suggesting that the dominant source of moisture for precipitation in the study area at the time of sampling was the Indian summer monsoon, rather than the winter westerlies, which have higher d (∼22‰) (Karim and Veizer, 2002). Compared to the Tso Moriri stream samples, headwater tributaries of the Indus River to the west in Pakistan show higher d values,

Conclusion

This study documented that evaporation comprises the majority of inflow to two deep lakes in the arid western Himalaya, indicating near complete basin closure to outflow. Two shallow lakes showed significant loss to outflow through groundwater, but large river systems did not show a strong signature of evaporative enrichment, and therefore likely have limited dependence on recharge from lakes experiencing large evaporation loss. Evaporatively enriched lakes may play a role in maintaining low

Acknowledgements

This project was funded by a grant from the Worldwide Fund for Nature-India (WWF-India). C.-T Lai was supported by the U.S. National Science Foundation, Division of Atmospheric and Geospace Sciences under Grant Number AGS-0956425. Many thanks to the WWF staff for help with the field campaign. Two anonymous reviewers provided valuable input that improved the analysis and manuscript.

References (41)

  • R. Ramesh et al.

    Stable isotope study of the Ganga (Ganges) river system

    J. Hydrol.

    (1992)
  • B. Zhang et al.

    Estimation and trend detection of water storage at Nam Co Lake, central Tibetan Plateau

    J. Hydrol.

    (2011)
  • S. Zhou et al.

    Water balance observations reveal significant subsurface water seepage from Lake Nam Co, south-central Tibetan Plateau

    J. Hydrol.

    (2013)
  • A. Zuber

    On the environmental isotope method for determining the water balance components of some lakes

    J. Hydrol.

    (1983)
  • T.N. Bagati et al.

    Quaternary basins of Ladakh and Lahaul-Spiti in northwestern Himalaya

    Curr. Sci.

    (1993)
  • B. Bookhagen et al.

    Toward a complete Himalayan hydrological budget: spatiotemporal distribution of snowmelt and rainfall and their impact on river discharge

    J. Geophys. Res.

    (2010)
  • B. Bookhagen et al.

    Late quaternary intensified monsoon phases control landscape evolution in the northwest Himalaya

    Geology

    (2005)
  • A. Bullock et al.

    The role of wetlands in the hydrological cycle

    Hydrol. Earth Syst. Sci.

    (2003)
  • F. Delclaux et al.

    Evaporation estimation on Lake Titicaca: a synthesis review and modelling

    Hydrol. Process.

    (2007)
  • T. Dincer

    Use of oxygen 18 and deuterium concentrations in water balance of lakes

    Water Resour. Res.

    (1968)
  • Cited by (36)

    • Spatial distribution of stable isotopes in surface water on the upper Indus River basin (UIRB): Implications for moisture source and paleoelevation reconstruction

      2022, Applied Geochemistry
      Citation Excerpt :

      Moreover, samples from the saline lake showed enriched isotopic composition (Fig. 2). The enriched isotopic composition reported here is fairly consistent with previous studies (Pande et al., 1994; Sharma et al., 2017; Biggs et al., 2015) for the saline lake samples. However, the low d-excess values in the lake are also due to the strong evaporation.

    • Impact of landscape dams on river water cycle in urban and peri-urban areas in the Shiyang River Basin: Evidence obtained from hydrogen and oxygen isotopes

      2021, Journal of Hydrology
      Citation Excerpt :

      The E/I value of the Shiyang River affected by human activities and drought was the largest compared with other research results. The E/I of the Yellow River source area in the Qinghai Tibet Plateau was 6% (Biggs et al., 2015), and the water yield was between 170 mm and 250 mm. The E/I of the lake area in Ontario, Canada was 9% (Gibson et al., 2017), and the water yield was 395 mm.

    • A case study using 2019 pre-monsoon snow and stream chemistry in the Khumbu region, Nepal

      2021, Science of the Total Environment
      Citation Excerpt :

      Similar δ18O values of KGM2 to the Khumbu Glacier ice core also suggest a stronger influence of snow/ice melt (Wood et al., 2020). More enriched δ18O values in the Imja Valley streams could be from evaporation losses (Biggs et al., 2015) due to their origin from a large glacial lake sourced from Imja Glacier melt. The Khumbu and Imja Valley samples exhibit similar concentrations for crustal elements, while the Pheriche location has higher concentrations for anthropogenically sourced chemistry (Pb, Cs, Bi, Cd, Ag, Cu).

    View all citing articles on Scopus
    View full text