The Indian and East Asian summer monsoons shape the melt and accumulation patterns of glaciers in High Mountain Asia in complex ways due to the interaction of persistent cloud cover, large temperature ranges, high atmospheric water content and high precipitation rates. Glacier energy- and mass-balance modelling using in situ measurements offers insights into the ways in which surface processes are shaped by climatic regimes. In this study, we use a full energy- and mass-balance model and seven on-glacier automatic weather station datasets from different parts of the Central and Eastern Himalaya to investigate how monsoon conditions influence the glacier surface energy and mass balance. In particular, we look at how debris-covered and debris-free glaciers respond differently to monsoonal conditions. The radiation budget primarily controls the melt of clean-ice glaciers, but turbulent fluxes play an important role in modulating the melt energy on debris-covered glaciers. The sensible heat flux decreases during core monsoon, but the latent heat flux cools the surface due to evaporation of liquid water. This interplay of radiative and turbulent fluxes causes debris-covered glacier melt rates to stay almost constant through the different phases of the monsoon. Ice melt under thin debris, on the other hand, is amplified by both the dark surface and the turbulent fluxes, which intensify melt during monsoon through surface heating and condensation. Pre-monsoon snow cover can considerably delay melt onset and have a strong impact on the seasonal mass balance. Intermittent monsoon snow cover lowers the melt rates at high elevation. This work is fundamental to the understanding of the present and future Himalayan cryosphere and water budget, while informing and motivating further glacier- and catchment-scale research using process-based models.
High Mountain Asia (HMA) holds the largest ice volume outside the polar regions
Accurate glacier mass-balance modelling is essential for assessing glacier meltwater contribution to mountain water resources, and for predicting future glacier states and catchment runoff. Physically based models of glacier energy and mass balance represent surface and subsurface energy fluxes using physical equations to calculate the energy available for melt and the glacier runoff. Summer-accumulation glaciers in HMA experience simultaneous accumulation and ablation. Using an energy-balance model,
The presence of debris cover, a widespread characteristic of HMA glaciers, (e.g.
Our main goal is to improve the understanding of monsoon controls on glaciers of various surface types in the Central and Eastern Himalaya. Applying the glacier energy- and mass-balance module of a land surface model suited to both debris-covered and clean-ice glaciers, and leveraging seven on-glacier automatic weather station (AWS) records from the region, we answer the following questions: (1) Which energy and mass fluxes dominate the seasonal mass balance of Himalayan glaciers? (2) How does debris cover modulate the energy balance in comparison with clean-ice surfaces? (3) How does the monsoon change the glacier surface energy balance? Answering these questions allows us to infer how these glaciers will respond to the possible future changes of the monsoons in the region. We apply the model at the point scale of individual AWSs, driven by high-quality in situ meteorological observations that guarantee accurate energy-balance simulations, not affected by extrapolation of the meteorological forcing. By identifying the key surface processes of glaciers and their dynamics under monsoonal conditions, this study promotes their appropriate representation in models of glacier mass balance and the hydrology of glacierised catchments.
In situ observations from seven on-glacier AWSs in different environments along the climatic gradient of the Himalaya were gathered and used for forcing and evaluation of the model (Fig.
North Changri Nup Glacier (hereafter Changri Nup Glacier) is a debris-covered valley glacier located in the Everest region in Nepal (Fig.
24K and Parlung No.4 glaciers are located on the southeastern Tibetan Plateau, feeding water into the upper Parlung Tsangpo, a major tributary to the Yarlung Tsangpo–Brahmaputra River. The summer climate is characterised by monsoonal air masses reaching the Gangrigabu mountain range from the south through the Yarlung Tsangpo Grand Canyon. 24K Glacier is an avalanche-fed valley glacier with a debris-covered tongue, located 24 km from the town of Bome
Hailuogou Glacier, the second-largest of our study sites (Fig.
We use the monthly averaged ERA5-Land re-analysis data
Characteristics of study sites, summarised (centre) in terms of elevation, mean measured ice melt rate, measured debris thickness and JJAS contribution to the ERA5-Land total annual (1981–2019) precipitation (monsoon index; MI). For each site, we also show glacier (bars in aqua) and debris (bars in olive) hypsometry, with area on the
Summary of available meteorological and ablation observations at each site, as well as each site's model period. Variables indicated by
Characteristics of the study sites. Planimetric glacier and debris surface areas, mean elevation, slope and aspect were calculated using the updated Randolph Glacier Inventory 6.0 by
We use the hydrological, snow and ice modules of the Tethys–Chloris (T&C) land surface model
For calculating the incoming energy with precipitation, rain is assumed to fall at air temperature (
The snow pack has a water-holding capacity
Over snow, debris and ice surfaces, the sensible energy flux is calculated as:
Correct estimates of the latent energy flux due to water phase changes at the surface are important for accurately modelling glacier melt, especially under moist conditions
The definition of the ground energy flux
The heat diffusion equation (Eq.
Note, that
Precipitation is partitioned into solid
The water content of ice is approximated with a linear reservoir model. The liquid water outflow is proportional to the ice pack water content
The water content of the snow pack
For supraglacial debris, both observations and methods for modelling its water content are lacking. We thus use a simplified scheme for moisture at the surface of the debris, in order to mimic the drying process of the debris surface: We assume debris to have a dynamic interception storage
The mass-balance calculation of snow and ice is somewhat similar, and therefore they are described together here. Calculations are performed for snow if there is snow precipitation during a time step or the modelled SWE at the surface is greater than zero. Net input of energy to the snow or ice pack will increase its temperature, and after the temperature has been raised to the melting point, additional energy inputs will result in melt. The change in the average temperature of the ice or snowpack d
The water equivalent mass of the snow/ice pack after melting WE
A major challenge in physically based mass-balance modelling of debris-covered glaciers is the selection of appropriate debris properties. In addition to the debris thickness, which was measured at the AWS location, values are needed for the thermal conductivity
Optimum debris parameters
We calculate the uncertainty associated with all energy- and mass-balance components by performing
Uncertainty ranges of parameters and input variables used for Monte Carlo runs. Where units are indicated with [–], the parameter or variable was perturbed by the fractional value shown.
The model accurately reproduces the measured surface height change (ablation and accumulation) at both debris-covered and clean-ice glaciers (Fig.
The ablation season average melt rates vary considerably across sites: The highest value of
The timing of snow cover is an important control both of the amounts and of the patterns of ice melt, as ice melt rates are close to zero during periods of snow cover. This becomes clear in Fig.
Melt rates of ice and snow (stacked) as weekly averages at each site. Vertical dotted lines indicate monsoon onset and end. Error bars depict the uncertainty (standard deviation) of the estimates. Melt rates are normalised to the mean of the ice melt over the entire period (value in the upper left of each panel).
The largest components in the energy balance are
Debris cover modulates the energy balance in several ways: With the albedo of the snow-free debris surface ranging between
During monsoonal conditions, increased cloudiness results in
We observe opposite changes in
Energy flux differences in the diurnal cycle (stacked) between pre-monsoon and monsoon. The direction of change is to be considered relative to the sign of the original flux. Positive and negative sign corresponds to energy added or removed from the glacier, respectively; grey background indicates debris-covered site, light blue indicates clean-ice sites and grey-blue indicates 1 cm debris site
Average
An interruption of the monsoon at 24K occurred in August 2016, possibly associated with an El Niño event
In contrast to the glaciers with thick debris, during the monsoon,
Our results are derived from simulations at one location (AWS) on each glacier. To understand how representative our results are of conditions across the glacier ablation zone at each site, and across the possible range of debris thicknesses in particular (Table S4), we conduct a sensitivity experiment to evaluate the transferability of our results across the glaciers' ablation areas (see detailed explanation in Supplement Sect. S5). This experiment shows that even accounting for the range of conditions across each glacier ablation area, the pattern of pre-monsoon to monsoon difference in flux components, and importantly the equalising effect on
Changes in the individual fluxes when moving from pre-monsoon to monsoon. Colour dots indicate `standard' runs with AWS site-specific conditions. Black bars indicate the uncertainty range on the standard runs. Grey indicates the sensitivity of flux changes (
Our results show the importance of the turbulent fluxes in the energy balance of debris-covered glaciers, their varying role as melt-enhancing or melt-reducing fluxes depending on the debris thickness, and how the monsoon modulates them.
To assess the controls on the turbulent fluxes, we regressed the modelled values of
Across the sites with thick debris, vpd has somewhat more power than
We apply our model in a systematic way to seven glaciers in a variety of environments in the Central and Eastern Himalaya. We force the model with in situ station data and constrain and evaluate it against observations of surface height change, lending great confidence to the energy flux components. Previous energy-balance studies in the region were limited to two
Previous energy-balance studies of debris-covered glaciers were limited to one or two study sites (e.g.
The ablation period occurs between April and November at all sites, and all sites are affected by the Indian and East Asian summer monsoons during this period (Figs. S2 to S8). A long-term average of
Overcast cloud cover, increased air temperatures and additional moisture modify the energy balance of debris-covered glaciers, to result in a melt-equalising effect between pre-monsoon and monsoon (Sect.
In contrast to debris-covered glaciers, when clean-ice glaciers are snow-free and the ice has been heated to the melting point, almost all net radiation goes into ice melt (Sect.
At the site with thin debris, we observe a melt-enhancing effect during monsoon conditions. The dark debris surface absorbs almost
Symbolic representation of changes in energy-balance components from pre-monsoon to monsoon. Triangles pointing down/up indicate a positive/negative flux with regard to our sign convention, where positive/negative means a flux towards/away from the surface, respectively. Red/blue indicate an increasing/decreasing value, respectively, of the flux when moving from pre-monsoon to monsoon. When signs switch, the underlying, empty triangles indicate the pre-monsoonal direction of the flux, while the overlying, coloured ones indicate the monsoonal flux.
Monsoon-influenced, summer-accumulation glaciers (such as Langtang, Lirung, Yala and Changri Nup) have been previously shown to be especially vulnerable to warming due to a decrease in accumulation and an enhancement of ablation due to reduced albedo
All future climate scenarios agree on continued warming during the 21st century over High Mountain Asia
The prospect of warmer temperatures together with increased precipitation would (1) cause a shift in the precipitation partition from snow to rain in the monsoon, resulting in snow cover shifting to higher elevations and increasing total melt; (2) potentially lead to an increase in early spring snowfall, which would delay the onset of ice melt; (3) increase the likelihood of ephemeral monsoonal snow cover but move it to higher elevations, thus leaving more of the lower ablation zones exposed; and (4) increase the wet-bulb temperature together with humidity to result in a further reduction of the solid fraction of precipitation during monsoon. Overall it is likely that glacier ablation zones will be exposed for longer periods under future monsoon climate due to a net decrease of the snow covered duration, with a resulting increase in total ablation. A lengthening of the monsoon into autumn, on the other hand,
The expected warmer and wetter monsoonal conditions, including increased cloudiness, will likely result in an overall increase of melt rates on clean-ice and glaciers with debris cover around or below the critical thickness. This is because (1) they are more directly controlled by net radiation (comprising both short- and long-wave fluxes), which is likely to increase in magnitude (Sect.
By applying an energy-balance model to seven sites across the Central and Eastern Himalaya, we have identified monsoon effects on the ablation season energy and mass balance of glaciers, common for the debris-covered and clean-ice glaciers studied here. A list of criteria used for choosing our modelling periods at each site is given in the Supplement Sect. S2. Applying these criteria, we chose one summer season record for each site, for which all required variables were available at a high level of data quality. As a result of this selection process, our analysis remained limited to one summer season at each site. Our work has also highlighted knowledge gaps which require further study: First, the influence of spring and monsoonal snow cover (its timing and amount) on the seasonal glacier mass balance is currently difficult to discern due to the paucity of multi-annual datasets in High Mountain Asia. Second, the timing and quantity of post-monsoon and winter precipitation influence the annual mass balance; however, even fewer datasets exist for the winter half-year in HMA, preventing a year-round analysis of similar detail. Third, all our sites are located in glacier ablation areas, and surface and energy mass fluxes will change with elevation. While we have tested how representative our point-scale results are for the entire ablation area of the glaciers considered, the response of glacier accumulation areas to monsoon remains to be investigated. Meteorological data from accumulation areas are scarce, however, limiting our current understanding. Future work should establish new year-round and multi-year records, including datasets from accumulation areas, in order to extend some of our findings. Future work could also target the spatial distribution of forcing data and parameters necessary to run energy-balance models at the glacier scale.
We model the energy and mass balance of seven glaciers in the Central and Eastern Himalaya at seven on-glacier weather stations. We find that:
At all sites, the largest mass loss component during the ablation season is ice melt, followed by snowmelt and sublimation, while the latter only plays a role at our highest sites and outside of the core monsoon. We find that the seasonal energy and mass balance is strongly controlled by variations of absorbed shortwave radiation, a result of the prevalence of spring snow cover and the occurrence of ephemeral monsoonal snow accumulation. Debris cover above the critical thickness returns most of the energy it absorbs back to the atmosphere via longwave emission and turbulent heat fluxes. While The response of the glacier mass and energy balance to the monsoon depends on the surface type: Melt rates tend to increase compared to the pre-monsoon at the clean-ice glaciers and the glacier with thin debris cover (with the exception of Yala), while they stay similar at the glaciers with thick debris cover. We attribute these differences to the role the turbulent fluxes play for each surface type. At the glaciers with thick debris cover, where the turbulent fluxes `work for' the glacier, evaporation of the additionally available moisture ( Given these findings, under projected future monsoonal conditions, namely warmer and possibly longer and wetter monsoons
All AWS datasets for the modelling periods considered in the analysis, together with ablation measurements, pre-processed forcing data, T&C model codes, outputs and scripts for analysing outputs are available under the following link:
The supplement related to this article is available online at:
StF, FP and EM designed the study. StF carried out the analysis with the help of CLF, MM and SiF. StF interpreted the results, created the figures and wrote the paper with the help of CLF, EM, MM, TES and FP. SiF, PW, WI, and QL reviewed the paper. WY and BD facilitated field data collection and provided parameterisations for albedo and precipitation phase. WY, PW and WI also contributed datasets.
The contact author has declared that neither they nor their co-authors have any competing interests.
Publisher's note: Copernicus Publications remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
We would like to thank our four anonymous reviewers, who's comments greatly helped to improve the paper. We would also like to thank Jakob Steiner and ICIMOD for hosting and contributing datasets. Our special thanks go to Marin Kneib for organising the field campaigns, and the Langtang and 24K field teams for making the data collection possible.
This research has been supported by the European Research Council, H2020 European Research Council (RAVEN (grant no. 772751)). The National Natural Science Foundation of China (41961134035) financially supported the data collection on 24K and Parlung No.4 glaciers.
This paper was edited by Marie Dumont and reviewed by four anonymous referees.