Periodic Reporting for period 2 - RAVEN (Rapid mass loss of debris covered glaciers in High Mountain Asia)
Reporting period: 2019-11-01 to 2021-04-30
One of the most important questions of climate change impact research today is how glaciers are responding to global warming. High Mountain Asia (HMA) glaciers represent the largest mass of ice outside the Polar Regions and provide water resources for millions of people in the headwaters of the major Asian rivers. Nevertheless, their historical changes and future trends are unclear, due to scarcity of observations and fundamental limitations in simulation models, which are not suited to represent the effects of rocky debris that covers the surface of many of the region’s glaciers.
RAVEN’s main goal is to understand past and future variations in HMA glaciers and water resources, with a special focus on the role that supraglacial debris forming on glacier surfaces plays in modulating glacier mass balance and runoff. Mantles of rock debris cover many glaciers across the spectrum of climate and morphological characteristics of HMA. However, the controls of debris distribution, characteristics and dynamics are not clear, and debris has been observed to form on both the large, dynamically active semi-arid glaciers of the Karakoram and the smaller, stagnating glaciers of the monsoon-dominated central Himalaya. Debris-covered glaciers’ role in catchment hydrology is also poorly understood, and they exhibit distinctive mass accumulation and melt processes that are not represented in models. They are fed by snow, ice and rock avalanches, rather than direct precipitation. The debris can greatly suppress melt rates, but debris mantles are often interrupted by hot spots of melt (ice cliffs and supraglacial ponds), leading to highly variable and atypical melt patterns. Finally, as they express mass loss through down-wasting on their tongues rather than terminus retreat, it is difficult to assess their changes from satellite images and ground measurements alone. Through these mechanisms, supra-glacial debris controls the precise response of HMA glaciers to a warming climate, but the little knowledge of its distribution and characteristics has to date prohibited an effective representation in models, and lead to uncertain regional projections of both glacier mass change and stream discharge response.
RAVEN seeks to overcome these barriers to elucidate the role of debris-covered glaciers in the water cycle of High Mountain Asia and establish how HMA glacier and runoff will evolve in the future. The project will determine the role that debris-covered glaciers play in the water cycle of HMA and represent them in assessments of future runoff using a multi-scale approach linking systematic in situ measurements with catchment and regional models. It will integrate scales (from the point scale of single ponds and cliffs forming on the glacier surface to the scale of an entire glacier and then catchment) and methodological approaches (numerical modelling and observational strategies) to understand which processes are necessary to adequately represent these glaciers in catchment, regional and global glacier models.
Our overall aim will be met by four specific objectives:
1. Characterise the occurrence of cliffs and ponds on debris-covered glaciers and understand their distribution, persistence and characteristics across a range of climates and glacial geomorphologies.
2. Advance the physical understanding of cliff and pond ablation and quantify their mass loss using novel energy balance models driven by field data and remote sensing observations.
3. Calculate the mass balance of debris-covered and debris-free glaciers by including cliff and pond ablation and evolution into an advanced glacier mass balance model.
4. Quantify future changes in glacier mass-balance and runoff of representative catchments of HMA.
RAVEN’s main goal is to understand past and future variations in HMA glaciers and water resources, with a special focus on the role that supraglacial debris forming on glacier surfaces plays in modulating glacier mass balance and runoff. Mantles of rock debris cover many glaciers across the spectrum of climate and morphological characteristics of HMA. However, the controls of debris distribution, characteristics and dynamics are not clear, and debris has been observed to form on both the large, dynamically active semi-arid glaciers of the Karakoram and the smaller, stagnating glaciers of the monsoon-dominated central Himalaya. Debris-covered glaciers’ role in catchment hydrology is also poorly understood, and they exhibit distinctive mass accumulation and melt processes that are not represented in models. They are fed by snow, ice and rock avalanches, rather than direct precipitation. The debris can greatly suppress melt rates, but debris mantles are often interrupted by hot spots of melt (ice cliffs and supraglacial ponds), leading to highly variable and atypical melt patterns. Finally, as they express mass loss through down-wasting on their tongues rather than terminus retreat, it is difficult to assess their changes from satellite images and ground measurements alone. Through these mechanisms, supra-glacial debris controls the precise response of HMA glaciers to a warming climate, but the little knowledge of its distribution and characteristics has to date prohibited an effective representation in models, and lead to uncertain regional projections of both glacier mass change and stream discharge response.
RAVEN seeks to overcome these barriers to elucidate the role of debris-covered glaciers in the water cycle of High Mountain Asia and establish how HMA glacier and runoff will evolve in the future. The project will determine the role that debris-covered glaciers play in the water cycle of HMA and represent them in assessments of future runoff using a multi-scale approach linking systematic in situ measurements with catchment and regional models. It will integrate scales (from the point scale of single ponds and cliffs forming on the glacier surface to the scale of an entire glacier and then catchment) and methodological approaches (numerical modelling and observational strategies) to understand which processes are necessary to adequately represent these glaciers in catchment, regional and global glacier models.
Our overall aim will be met by four specific objectives:
1. Characterise the occurrence of cliffs and ponds on debris-covered glaciers and understand their distribution, persistence and characteristics across a range of climates and glacial geomorphologies.
2. Advance the physical understanding of cliff and pond ablation and quantify their mass loss using novel energy balance models driven by field data and remote sensing observations.
3. Calculate the mass balance of debris-covered and debris-free glaciers by including cliff and pond ablation and evolution into an advanced glacier mass balance model.
4. Quantify future changes in glacier mass-balance and runoff of representative catchments of HMA.
The project has made significant progress towards reaching its objectives in the first 30 months. Field expeditions to the Himalayas (Nepal and Tibet) were organized and key datasets regarding glacier properties, snow, hydrology and meteorology were collected successfully. Important scientific advances have been made, which we elaborate here grouped by each work package (corresponding to the four specific objectives above).
WP1: Cliff and pond distribution, persistence and evolution from satellite and field based observations
At the very large scale, we have completed three distinct pieces of ground-breaking, high impact research to assess debris covered glaciers using satellite images. We completed a global assessment of debris cover ‘stage of development’, published in Nature Geoscience, providing a framework to understand different glaciers’ distinct expressions of debris cover, ice cliffs, and supraglacial ponds. A major breakthrough was the development of an automated workflow to correct satellite geodetic imagery for observed ice motion in order to resolve glacier altitudinal mass balance, which we applied to resolve glacier altitudinal mass balance for 5527 glaciers across High Mountain Asia, revealing the distinctive mass balance patterns of debris-covered glaciers across the region (in review for Nature Communications). We have leveraged the two datasets above to develop a novel method to determine supraglacial debris thickness and mean headwall erosion rates across High Mountain Asia, which we have used to demonstrate the climatic and topographic controls of glaciers’ debris cover.
We have collected and analyzed extensive high-quality data of supraglacial ice cliffs and ponds across glaciers to understand their spatial distributions and variability. We completed the acquisition and processing of Pleiades stereo satellite imagery and UAV multi-view stereo imagery for 24K and Langtang sites, and developed a new method for automated mapping of ice cliffs on debris-covered glaciers. We have analysed high-resolution multispectral satellite imagery to produce a 2009-2019 time-series of ice cliff and supraglacial pond coverage for the 4 RAVEN sites, which have highlighted the stochastic and externally-driven cliff and pond variability.
Finally, we have complemented the remote sensing observations with systematic in situ measurements. We configured and tested hydrometeorological equipment for monitoring debris-covered glaciers. We developed novel time-lapse camera arrays for monitoring of ice cliffs and supraglacial ponds between site visits. We completed 2 field seasons each for Langtang (Nepal) and 24K (China) field sites in 2019. In 2020, we were unable to visit intended RAVEN sites due to COVID-19, but completed 1 low-cost field season studying ice cliffs on Zmuttgletscher (CH).
WP2. Physically based modelling of cliff and pond ablation and dynamics
At the 30-month point, RAVEN has made considerable progress towards its objective of modelling cliff and pond energy balance. We completed the first catchment-wide analysis of supraglacial pond-associated ablation for the Langtang catchment. We similarly completed the first catchment-wide analysis of cliff-associated ablation. We have tested the suitability of available ice cliff models at other sites in HMA, and used available models of cliff, pond, and sub-debris melt to conduct a global assessment of these surfaces’ melt rates relative to a clean ice glacier. Finally, we have developed a stochastic model of cliff and pond population dynamics and melt for glacier-scale, long-term applications.
WP3. Modelling the mass balance of debris-covered and debris-free glaciers
Our project has made several key advances in modelling the mass balance of glaciers in High Mountain Asia. We have implemented supraglacial debris within a novel, hyper-resolution land surface model (Tethys Chloris) suited for high mountain catchments, which we intend to use in the remaining of the project. We have evaluated the model’s performance at the RAVEN sites and used the model to examine the influence of the monsoon on the energy balance of debris-covered and clean ice glaciers across HMA. We have also led the first debris-covered glacier melt model intercomparison, which has highlighted the strengths and weaknesses of diverse methods to calculate sub-debris melt and indicated the most suitable melt modelling approach for RAVEN.
The implementation of supraglacial debris within Tethy-Chloris is a key advance for glacier mass balance in High Mountain Asia, so we have carefully assessed the appropriate meteorological forcing and validation data for this new generation of mass balance model. To constrain high-altitude glacier accumulation, we have measured high-altitude accumulation rates using a manual corer and borehole camera, representing extremely rare data. We have also demonstrated the utility of glaciological, hydrological, and meteorological observations in constraining a model of long-term (1974-2017) glacier mass balance and catchment hydrology in SE Tibet. Finally, we have evaluated the newest available climate reanalysis products and downscaling methods for high-elevation, glacierized catchments using an extensive network of weather stations across the region, as well as unique high-elevation datasets newly available from ourselves and our partners.
WP4. Future changes in glacier mass balance and runoff
In preparation for this work package, which integrates the new understanding from the local and regional observations into our new glacio-hydrological model, our work to date has centered on the compilation of input datasets for each study site. We have also developed codes for non-linear bias correction of climate data and assembled the most appropriate routines to best use the large amount of coarse-resolution climate data as forcing to our hyper-resolution model.
WP1: Cliff and pond distribution, persistence and evolution from satellite and field based observations
At the very large scale, we have completed three distinct pieces of ground-breaking, high impact research to assess debris covered glaciers using satellite images. We completed a global assessment of debris cover ‘stage of development’, published in Nature Geoscience, providing a framework to understand different glaciers’ distinct expressions of debris cover, ice cliffs, and supraglacial ponds. A major breakthrough was the development of an automated workflow to correct satellite geodetic imagery for observed ice motion in order to resolve glacier altitudinal mass balance, which we applied to resolve glacier altitudinal mass balance for 5527 glaciers across High Mountain Asia, revealing the distinctive mass balance patterns of debris-covered glaciers across the region (in review for Nature Communications). We have leveraged the two datasets above to develop a novel method to determine supraglacial debris thickness and mean headwall erosion rates across High Mountain Asia, which we have used to demonstrate the climatic and topographic controls of glaciers’ debris cover.
We have collected and analyzed extensive high-quality data of supraglacial ice cliffs and ponds across glaciers to understand their spatial distributions and variability. We completed the acquisition and processing of Pleiades stereo satellite imagery and UAV multi-view stereo imagery for 24K and Langtang sites, and developed a new method for automated mapping of ice cliffs on debris-covered glaciers. We have analysed high-resolution multispectral satellite imagery to produce a 2009-2019 time-series of ice cliff and supraglacial pond coverage for the 4 RAVEN sites, which have highlighted the stochastic and externally-driven cliff and pond variability.
Finally, we have complemented the remote sensing observations with systematic in situ measurements. We configured and tested hydrometeorological equipment for monitoring debris-covered glaciers. We developed novel time-lapse camera arrays for monitoring of ice cliffs and supraglacial ponds between site visits. We completed 2 field seasons each for Langtang (Nepal) and 24K (China) field sites in 2019. In 2020, we were unable to visit intended RAVEN sites due to COVID-19, but completed 1 low-cost field season studying ice cliffs on Zmuttgletscher (CH).
WP2. Physically based modelling of cliff and pond ablation and dynamics
At the 30-month point, RAVEN has made considerable progress towards its objective of modelling cliff and pond energy balance. We completed the first catchment-wide analysis of supraglacial pond-associated ablation for the Langtang catchment. We similarly completed the first catchment-wide analysis of cliff-associated ablation. We have tested the suitability of available ice cliff models at other sites in HMA, and used available models of cliff, pond, and sub-debris melt to conduct a global assessment of these surfaces’ melt rates relative to a clean ice glacier. Finally, we have developed a stochastic model of cliff and pond population dynamics and melt for glacier-scale, long-term applications.
WP3. Modelling the mass balance of debris-covered and debris-free glaciers
Our project has made several key advances in modelling the mass balance of glaciers in High Mountain Asia. We have implemented supraglacial debris within a novel, hyper-resolution land surface model (Tethys Chloris) suited for high mountain catchments, which we intend to use in the remaining of the project. We have evaluated the model’s performance at the RAVEN sites and used the model to examine the influence of the monsoon on the energy balance of debris-covered and clean ice glaciers across HMA. We have also led the first debris-covered glacier melt model intercomparison, which has highlighted the strengths and weaknesses of diverse methods to calculate sub-debris melt and indicated the most suitable melt modelling approach for RAVEN.
The implementation of supraglacial debris within Tethy-Chloris is a key advance for glacier mass balance in High Mountain Asia, so we have carefully assessed the appropriate meteorological forcing and validation data for this new generation of mass balance model. To constrain high-altitude glacier accumulation, we have measured high-altitude accumulation rates using a manual corer and borehole camera, representing extremely rare data. We have also demonstrated the utility of glaciological, hydrological, and meteorological observations in constraining a model of long-term (1974-2017) glacier mass balance and catchment hydrology in SE Tibet. Finally, we have evaluated the newest available climate reanalysis products and downscaling methods for high-elevation, glacierized catchments using an extensive network of weather stations across the region, as well as unique high-elevation datasets newly available from ourselves and our partners.
WP4. Future changes in glacier mass balance and runoff
In preparation for this work package, which integrates the new understanding from the local and regional observations into our new glacio-hydrological model, our work to date has centered on the compilation of input datasets for each study site. We have also developed codes for non-linear bias correction of climate data and assembled the most appropriate routines to best use the large amount of coarse-resolution climate data as forcing to our hyper-resolution model.
The project has already resulted in a number of key advances beyond the current state of the art and several groundbreaking insights in High Mountain Asia, in particular regarding the stage of development of supraglacial debris cover in the region, as well as an unprecedented examination of the current health and long-term sustainability of the region’s glaciers.
WP1: Cliff and pond distribution, persistence and evolution from satellite and field based observations
Our development of a new, transferable method for automated mapping of ice cliffs on debris-covered glaciers opens the door to systematic, robust mapping of both ice cliffs and supraglacial ponds across High Mountain Asia. By the end of the project we will complete the analyses of seasonal and interannual ice cliff and supraglacial pond dynamics at the four study catchments and a large-scale assessment of supraglacial ice cliff and pond extent and their topographic and glaciological controls.
At the large scale, we have generated a global assessment of glacier debris cover stage of development, which developed a very novel understanding of the stage of evolution of supraglacial debris over all glaciers worldwide. It demonstrated that debris-covered glaciers in High Mountain Asia are at an advanced state of decay, and allowed understanding the conditions over which ice cliffs and ponds can form.
Finally, a major breakthrough has been a new systematic analysis of glacier surface mass balance across High Mountain Asia in the first large-scale application of the continuity equation, revealing the poor glacier health in the region in unprecedented detail. Additional expected large-scale results before the end of the project include an analysis of supraglacial debris thickness and implied debris supply rates across High Mountain Asia and an examination of the importance of sub-debris melt into rivers across High Mountain Asia.
WP2. Physically based modelling of cliff and pond ablation and dynamics
Our work so far has provided the first understanding of the melt contribution of ice cliffs and supraglacial ponds at the catchment scale. We have also completed a global comparison of the energy balance of ice cliffs, supraglacial ponds, and debris-covered ice, extending understanding of associated melt rates beyond High Mountain Asia. Before the end of the project, we expect several key advances beyond the current understanding of these features. A new stochastic population dynamics model of ice cliff and supraglacial pond variations and melt will for the first time disentangle these features’ internal dynamics from the influence of external drivers. New datasets already collected from time-lapse camera arrays at the Langtang, 24K, and Zmuttgletscher sites will provide insight into the subseasonal processes and dynamics of ice cliffs, supraglacial hydrology, and debris mobilization rates, understanding that is currently lacking. These individual advances will be drawn together into a physically-grounded yet computationally-efficient model of ice cliff and supraglacial pond influences to be integrated into glacier models.
WP3. Modelling the mass balance of debris-covered and debris-free glaciers
The completion of the first debris-covered glacier melt model intercomparison has been a major achievement by the project team, and has resulted in a comprehensive understanding of the strengths and weaknesses of individual models. We have also performed the first multi-site investigation of the relative influence of the monsoon, debris thickness, and elevation on the energy balance of debris-covered and clean ice glaciers in HMA. The use and development of a hyper-resolution land surface model that represents the mountain cryosphere, hydrosphere and biosphere in a much more advanced and substantially distinct manner from traditional glacio-hydrological models represents a key element of novelty of the project. By the end of the project, we will quantify the importance of debris-covered glaciers for catchment water balance at the four study sites based on a hyper-resolution land surface model and provide a first analysis of blue-green water interactions in high-mountain Asian catchments.
WP4. Future changes in glacier mass balance and runoff
By the end of the project, we will produce novel long term projections of glacier and streamflow change at four catchments in High Mountain Asia that for the first time: i) describe all components of the cryosphere and hydrosphere using physics-based equations, and do not reply on empirical simplifications; ii) in particular, include the key role of debris in modulating glacier response at short and long time scales, with the potential of importantly altering the trajectory of cryospheric response; and iii) describe the complex interrelationship between cryosphere, hydrosphere and biosphere, and will allow for feedback between processes to modulate the catchment response to a warming climate. Using the hyper-resolution land surface model, we will elucidate the importance of debris, ice cliffs and supraglacial ponds to future glacier mass balance and change across High Mountain Asia, as well as the changing roles of vegetation phenology and glacier seasonal melt in controlling high-mountain catchment water budgets.
WP1: Cliff and pond distribution, persistence and evolution from satellite and field based observations
Our development of a new, transferable method for automated mapping of ice cliffs on debris-covered glaciers opens the door to systematic, robust mapping of both ice cliffs and supraglacial ponds across High Mountain Asia. By the end of the project we will complete the analyses of seasonal and interannual ice cliff and supraglacial pond dynamics at the four study catchments and a large-scale assessment of supraglacial ice cliff and pond extent and their topographic and glaciological controls.
At the large scale, we have generated a global assessment of glacier debris cover stage of development, which developed a very novel understanding of the stage of evolution of supraglacial debris over all glaciers worldwide. It demonstrated that debris-covered glaciers in High Mountain Asia are at an advanced state of decay, and allowed understanding the conditions over which ice cliffs and ponds can form.
Finally, a major breakthrough has been a new systematic analysis of glacier surface mass balance across High Mountain Asia in the first large-scale application of the continuity equation, revealing the poor glacier health in the region in unprecedented detail. Additional expected large-scale results before the end of the project include an analysis of supraglacial debris thickness and implied debris supply rates across High Mountain Asia and an examination of the importance of sub-debris melt into rivers across High Mountain Asia.
WP2. Physically based modelling of cliff and pond ablation and dynamics
Our work so far has provided the first understanding of the melt contribution of ice cliffs and supraglacial ponds at the catchment scale. We have also completed a global comparison of the energy balance of ice cliffs, supraglacial ponds, and debris-covered ice, extending understanding of associated melt rates beyond High Mountain Asia. Before the end of the project, we expect several key advances beyond the current understanding of these features. A new stochastic population dynamics model of ice cliff and supraglacial pond variations and melt will for the first time disentangle these features’ internal dynamics from the influence of external drivers. New datasets already collected from time-lapse camera arrays at the Langtang, 24K, and Zmuttgletscher sites will provide insight into the subseasonal processes and dynamics of ice cliffs, supraglacial hydrology, and debris mobilization rates, understanding that is currently lacking. These individual advances will be drawn together into a physically-grounded yet computationally-efficient model of ice cliff and supraglacial pond influences to be integrated into glacier models.
WP3. Modelling the mass balance of debris-covered and debris-free glaciers
The completion of the first debris-covered glacier melt model intercomparison has been a major achievement by the project team, and has resulted in a comprehensive understanding of the strengths and weaknesses of individual models. We have also performed the first multi-site investigation of the relative influence of the monsoon, debris thickness, and elevation on the energy balance of debris-covered and clean ice glaciers in HMA. The use and development of a hyper-resolution land surface model that represents the mountain cryosphere, hydrosphere and biosphere in a much more advanced and substantially distinct manner from traditional glacio-hydrological models represents a key element of novelty of the project. By the end of the project, we will quantify the importance of debris-covered glaciers for catchment water balance at the four study sites based on a hyper-resolution land surface model and provide a first analysis of blue-green water interactions in high-mountain Asian catchments.
WP4. Future changes in glacier mass balance and runoff
By the end of the project, we will produce novel long term projections of glacier and streamflow change at four catchments in High Mountain Asia that for the first time: i) describe all components of the cryosphere and hydrosphere using physics-based equations, and do not reply on empirical simplifications; ii) in particular, include the key role of debris in modulating glacier response at short and long time scales, with the potential of importantly altering the trajectory of cryospheric response; and iii) describe the complex interrelationship between cryosphere, hydrosphere and biosphere, and will allow for feedback between processes to modulate the catchment response to a warming climate. Using the hyper-resolution land surface model, we will elucidate the importance of debris, ice cliffs and supraglacial ponds to future glacier mass balance and change across High Mountain Asia, as well as the changing roles of vegetation phenology and glacier seasonal melt in controlling high-mountain catchment water budgets.