Author Archives: Philip Kraaijenbrink

Socio-economic development key driver future South Asian water gap

The recent United Nations Climate Change Conference COP24 held in Katowice, Poland once more demonstrated the world’s climate change concerns. In the Indus, Ganges, and Brahmaputra river basins, a global climate change hotspot and home for about 900 million people, these concerns are pressing, since the river systems provide water resources for the important agricultural, domestic, and industrial sectors that serve these people. Melt water from glaciers and snow feed the headwaters of these rivers and are strongly influenced by rising temperatures. In addition, the monsoon and its dynamics, which determine the regional hydrology, are expected to change. Moreover, strong socio-economic developments and a rapid and continuous population growth will result in tremendous increases in water demand and cause pressure on water resources. It is therefore very likely that a water gap will develop in the future.

A new study published (open access) in Hydrology and Earth System Sciences led by René assesses the combined impacts of climate change and socio-economic developments on the future “blue” water gap in the Indus, Ganges, and Brahmaputra river basins until the end of the 21st century. In this joint effort by FutureWater, Utrecht University, Wageningen Environmental Research and ICIMOD a hydrological model that simulates future changes in the upstream water reserves (SPHY) is coupled with a hydrology and crop production model that simulates future changes in the downstream water balance (LPJmL). The models were forced with the latest climate change projections and socio-economic scenarios.

The findings of this study indicate that the surface water availability will increase, which can mainly be attributed to increases in monsoon precipitation. Besides the increases in surface water availability, water consumption by irrigation will most likely decline due to shorter growing seasons that emerge from temperature increases, and a shift from blue water irrigation to green water/rainfed irrigation due to increases in precipitation. However, this increase in water availability cannot outweigh the strong increases in water demand that are associated with the strong socio-economic development, and will thus likely lead to a substantial increase in the water gap with 7% and 14% in the Indus and Ganges river basins, respectively, during the 21st century. This implies the importance of robust adaptation strategies to cope with future water shortages in the region.


Maps showing the annual groundwater depletion for the reference period (a) and the projected changes in groundwater depletion for RCP4.5 (b), RCP8.5 (c), RCP4.5 – SSP1 (d), and RCP8.5 – SSP3 (e). The projected changes are given for the end of the 21st century. Green indicates less depletion and red indicate more depletion.

Spatial precipitation patterns resolved with atmospheric modelling

Our new open access study, led by Pleun, shows the importance of subkilometer atmospheric modelling and correct surface boundary conditions in areas with complex topography to accurately estimate catchment-scale meteorological variability.

Frequently used gridded meteorological datasets poorly represent precipitation in the Himalayas because of their relatively low spatial resolution and the associated representation of the complex topography. Dynamical downscaling using high-resolution atmospheric models may improve the accuracy and quality of the precipitation fields. Therefore, we have used the Weather Research and Forecasting (WRF) Model to determine the resolution that is required to most accurately simulate monsoon and winter precipitation, 2-m temperature, and wind fields in the Nepalese Himalayas.

Results show that a high resolution of 500 m is computationally still feasible and provides the best match with the observations, gives the most plausible spatial distribution of precipitation, and improves the quality of the wind and temperature fields. Our findings suggest that, in combination with future improvements to atmospheric models for applications in complex terrain, subkilometer grid spacing may resolve catchment-scale meteorological variability more accurately. This will improve our capabilities to study glacio-hydrological changes at catchment and larger scale. Future modeling studies of High Mountain Asia should consider subkilometer grids to accurately estimate local meteorological variability.

P.N.J. Bonekamp, E. Collier, W.W. Immerzeel (2018), The Impact of Spatial Resolution, Land Use, and Spinup Time on Resolving Spatial Precipitation Patterns in the Himalayas, Journal of Hydrometeorology, 19, 1565-1581.


Model output compared to the observations for the summer (left panels) and winter period (right panels).


The effect of spatial resolution for the simulated winter and summer period.

PhD degree for Philip Kraaijenbrink

Wednesday September 19th Philip successfully defended his PhD during the formal Utrecht University defense ceremony and he may now call himself Dr. Kraaijenbrink.

Philip has received his PhD with distinction for his thesis titled High-resolution insights into the dynamics of Himalayan debris-covered glaciers in which he improved our understanding of debris-covered glaciers by studying them in detail with unmanned aerial vehicles and modelling. The thesis is available for open access download via the Utrecht University library.

Philip was supervised by Prof. Steven de Jong, Dr. Walter Immerzeel and Dr. Joseph Shea. The external defense committee was formed by Prof. Andreas Kääb, Prof. Etienne Berthier, Prof. Koji Fujita, Prof. Michiel van den Broeke and Dr. Francesca Pellicciotti.


Prof. dr. Steven de Jong provides Philip with his degree in the formal Utrecht University tradition.

Snow sublimation on a Himalayan glacier

The eddy covariance tower on Yala Glacier after installation in October 2016 (Photo: Walter Immerzeel).

Our new study, which was led by Emmy and published in Frontiers in Earth Science, shows that snow sublimation should no longer be ignored in future hydrological and mass balance studies in the Himalaya. We assessed the importance of snow sublimation to the water and mass budget of Yala Glacier in the Langtang Valley, Nepalese Himalaya.

From a hydrological and glaciological perspective, snow sublimation is a loss of water from the snowpack to the atmosphere. So far, snow sublimation has remained unquantified in the Himalaya, prohibiting a full understanding of the water balance and glacier mass balance. Hence, we measured surface latent heat fluxes with an eddy covariance system on Yala Glacier (5350 m a.s.l) to quantify the role snow sublimation plays in the water and glacier mass budget.

The observed sublimation is 32 mm for a 32-day period from October to November 2016, which is high compared to observations in other regions in the world. The bulk-aerodynamic method was used to estimate cumulative sublimation and evaporation at the location of the eddy covariance system for the 2016–2017 winter season, which is 125 and 9 mm respectively. This is equivalent to 21% of the annual snowfall.

A combination of meteorological observations and WRF simulations were used to estimate the spatial variability in sublimation. These simulations reveal that sublimation is primarily controlled by wind speed. The daily cumulative sublimation is a factor 1.7 higher at the ridge of Yala Glacier, which is wind-exposed, compared the location of the eddy covariance system. This is a considerable loss of water and illustrates the importance and need to account for sublimation in future studies in the Himalaya.

This work quantifies surface sublimation only. However, sublimation may be enhanced under conditions with wind-induced snow transport. Therefore, future research will focus on including this component to fully assess the importance of snow sublimation in the high-altitude water cycle.


Open access article:
Stigter, E. E., Litt, M., Steiner, J. F., Bonekamp, P. N. J., Shea, J. M., Bierkens, M. F., & Immerzeel, W. W. (2018). The importance of snow sublimation on a Himalayan glacier. Frontiers in Earth Science6 (108), 1-16.


Scatter plots of meteorological variables against sublimation rate, observed at AWS Yala Glacier. The color of the data points refers to the observed wind speed. Results show that vapor pressure deficit and wind speed are the best sublimation predictors.

Mapping surface temperatures using a UAV

A new study led by Philip presents a method to map surface temperatures of a debris-covered glacier with an unmanned aerial vehicle (UAV), which has potential to study melt processes of such glaciers. It was published open access today in Frontiers of Earth Sciences.

In the paper we map surface temperatures of Lirung Glacier in three flights on a morning in May 2016. We present a methodology to georeference and process the acquired thermal imagery, and correct for emissivity and sensor bias. Derived UAV surface temperatures are compared with distributed simultaneous ground-based temperature measurements and with Landsat 8 thermal satellite imagery.

Surface temperatures vary greatly both spatially and temporally and have a large range of 50 °C over course of the morning. Statistical analysis shows that the variability is largely independent of incoming radiation and topography, and that much of the signal in surface temperature originates from variation in properties of the debris. Future research of surface melt processes can utilize this data to further unravel heterogeneous melt patterns on debris-covered glaciers.


Kraaijenbrink, P. D. A., J. M. Shea, M. Litt, J. F. Steiner, D. Treichler, I. Koch, and W. W. Immerzeel (2018)
Mapping surface temperatures on a debris-covered glacier with an unmanned aerial vehicle
Frontiers in Earth Science, 6(64), 1–19.


Surface temperature orthomosaics of the three UAV flights on 1 May 2016 (A–C; 06:45, 09:20, and 10:35) and the brightness temperature of the Landsat 8 band 10 on 2 May 2016 at 10:32.


Comparison of the average warming rate (f) over the surveyed glacier surface area with five different DEM derivatives: aspect (a,g), slope (b,h), upstream area (c,i), relative local elevation (d,j), and mean incoming shortwave radiation (e,k). The relative importance of each variable as a predictor in a random forest regression is shown in (l)


Animation of ground-based thermal imaging of an ice cliff and its surroundings, performed synchonous to the UAV flights.


1 2 3 8