Glacial Diminishment of the Gries Glacier, Switzerland
- Krista Emery
- Jul 24, 2019
- 13 min read
Updated: Apr 28, 2023
Contents
Study Question and Significance
Background
Alps Climate
Glacier Basics
Positive Feedback Mechanisms
Methods
Remote Sensing
NDSI
Change Detection
Change Detection and Binary Change Detection
Binary Change
Results
NDSI
Change Detection
Binary Change Analysis and Quantification of Decline
Discussion
Limitations
Conclusion
Recommendations and Further Analysis
Study Question and Significance
This project aims to quantify the amounts of change exhibited by the Gries glacier in the Swiss Alps between 1985, 2000, and 2015. The Swiss Alps are a long chain of high-altitude mountainous terrain. The geographic positioning of these mountains in the midlatitudes combined with frequent precipitation, providing glaciers from the last Ice Age an environment to continue existing to this day. (World Glacier Monitoring Service 2009, 12) Because glaciers are such sensitive indicators of climate changes, it reinforces the importance these types of studies for present and future global climate studies. Global climate studies are necessary for major changes in global climate responses. (Levermann, et al. 2012, 864) The location of the glacier in relativity to the study area are indicated in figure 1. In the analysis of the Gries glacier outlined in this study, methods will include utilizing remote sensing imagery to derive landcover data.
Background
Most people know that changes are occurring on our planet in terms such as climate change and global warming. Some may say something along the lines of "the ice sheets are melting" when referring to sea level rise and evidence of a changing climate. However, it's not just ice sheets like Greenland and the Antarctic ice sheets that are suffering under the stressors of climate change.
Alps Climate
Over the past 30 years, temperatures have been recorded to be increasing and significant precipitation becoming less and less prominent in the Alps region in more recent times. The graph in figure two shows the overall trends of air temperature and precipitation in the Gries area.
Based on previous research, the selected area for this study was decided to be the Gries glacier. (Huss, Bauder, et al. 2008) Figures 3, 4, and 5 show previous record of mass balances in the Alps region, and overall trends. Shown in figure 4, 2003 was an exceptional year in terms of annual mass loss due to a heat wave that resulted in many casualties throughout Europe.
Based on previous research, the selected area for this study was decided to be the Gries glacier. (Huss, Bauder, et al. 2008) Figures 3, 4, and 5 show previous record of mass balances in the Alps region, and overall trends.
Shown in figure 4, 2003 was an exceptional year in terms of annual mass loss due to a heat wave that resulted in many casualties throughout Europe.
Glaciers in mountain ranges like the Alps provide important freshwater resources for communities and to contribute to the Rhine and Rhône Rivers. The lack of glaciers in the Alps would result in a drought throughout Europe. (Levermann, et al. 2012, 862-3; Rebetez 1996)
Glacier Basics
A glacier is defined as a flowing mass of surface ice that moves downhill due to forces of gravity, constrained by its internal stress, and friction at its bases and sides. Because alpine glaciers exist in mountainous areas with high altitudes and don't experience calving or surging like ice sheets and other forms of perennial ice, they are great indicators of a changing climate. Climate forcing is a reaction that climatic indicator features such as glaciers have to climate variables such as solar radiation, air temperature, precipitation, wind direction/speed, atmospheric greenhouse gases, or cloudiness. (WGMS 2009, 10; Levermann et al., 2012, 864)
Fluctuations exhibited by glaciers can be quantified indirectly through the change in length, or advance and retreat. Indirect reactions are delayed and filtered indicators of climate change trends. Fluctuations that can be quantified directly are the glacier mass balance, or changes in thickness or volume of ice contained in a glacial body. The mass balance is not a delayed response to annual atmospheric factors, unlike the horizontal length changes. (WGMS, 12) Figure 5 shows the generalized seasonal variation of mass balance that the area studied in this paper follows.
However, glacial changes are unique to each individual glacier. Balances are influenced by regional climate variability as well as local topographic effects like elevation, incoming net radiation, or accumulation patterns. These topographic effects are so significant that two neighboring glaciers may experience significantly different responses at the same time period. Surface mass balance is strongly influenced by debris cover on the ice as well. (WGMS, 12; Vincent, et al. 2017, 1380)
The Gries glacier has had consistent measurements over the past few decades, allowing for reliable change data to be viewed and analyzed. Figure 6 shows the differences in mass balance records for the Gries glacier on fifteen-year intervals from 1985 to 2015.
2015 has been the leader in terms of yearly mass loss and summer mass loss compared to the two previous time steps in question. There is a trend that shows a huge decline in mass balance over time. (VAW/ETH Zuerich) Articles by Zemp and Fischer gave insight on how to complete mass balance computations, as well as raw data from the glacier. (Zemp, et al. 2008; Fischer, Huss and Hoelzle 2015)
Temperate alpine glaciers experience ablation and accumulation in response to changes in the mentioned variables. Temperate glaciers are formed and maintained by accumulation of snow at the higher altitudes of the feature during humid months, or winter. The snow densifies and transforms into perennial firn, then the pressure and friction push the air out, forcing the snow to become ice. Different locations of glaciers result in different responses and characteristics, making glaciers a topic of wide variety. Because of this variety, analytical or numerical modelling is required to quantify each variable, its topographic effect, and climate parameters. (WGMS, 12)
The accumulation is offset by the melting, or ablation in lower altitudes of the feature. Discharge into lakes or to the sea are also ways in which meltwater can exit the feature. The area between the ablation and accumulation is the equilibrium line. This represents the balance between loss and gain is exactly zero. The equilibrium line can move based on the climate variables influencing its accumulation or ablation. The distribution of the glacial ice is a function of mean annual air temperature and annual precipitation sums. (WGMS, 12)
Positive Feedback Mechanisms
An important mechanism to mention is the self-amplification processes associated with climate change and melting snow. Effects of the climate variables on the glacier are influenced by the ice-albedo effect, or the positive feedback mechanism related to light reflecting off lightly or darkly colored backgrounds. When light hits ice, it increases the reflectivity of the sunlight and bounces off as opposed to a darker surface like dirt or rock where it is absorbed. The presence of glaciers promotes the presence of other glaciers based on the principles of the ice albedo effect reflecting rather than absorbing the solar radiation striking its surface. (Naegeli and Huss 2017; Kääb, Machguth and Paul 2005)
Permafrost and ice melting increase the presence of carbon dioxide and methane in the atmosphere because it had been sealed into the ice for thousands of years. Therefore, the more that climate is changing and causing permafrost and ice to melt, the faster that the climate will continue to change. (Levermann, et al. 2012, 848-63)
Methods
The methods for this analysis include remote sensing imagery preprocessing, NDSI, change detection, and binary change detection. My methods were influenced by dozens of articles on remote sensing and glacial measurements, but several provided a good basis on which I could base my analysis.
Remote Sensing
Remote sensing imagery is used to identify areas of significant changes in the Gries glacier and attempt to acquire the amount of change in an understandable fashion. My satellite data used for remote sensing purposes came from two different satellites: Landsat 5 Thematic Mapper and Landsat 8 Operational Land Imager. This is due to the differences in orbit dates, as Landsat 5 was only available until 2011, and my study time period was outside of that range. The resolution of the area for each image is 30 x 30 m, which is used when converting change detection pixel values to real-world units. Tables 1 and 2 below show the image dates, resolutions, band designations, and wavelengths associated with each sensor used in the analysis.
Prior to any NDSI or change detection analysis, each image had to go through preprocessing. Bands were selected, stacked, orthorectified, and subsetted. For the Landsat 5 TM images the band selection was 1-5 and 7. The Landsat 8 OLI image consisted of a stack with bands 2-7 to correspond with the other sensor’s bands. At least 20 ground control points were used to orthorectify the images to ensure that they were correctly aligned for comparative analysis. The area of study was cropped through a simple subset for each image by the same parameters.
NDSI
NDSI, or the normalized difference snow index was a method utilized in Silverio and Jaquet 2005 as well as Nijhawan et al. 2014 to highlight differences between ice-snow areas from non-ice-snow areas.
The NDSI is a remote sensing index in which spectral bands that show reflectance of various surfaces from space are categorized and isolated. In this ratio of bands, it takes advantage of the differences between snow in short-wave infrared and visible (red, green, blue) portions of the light spectrum against other features in my study area. In the visible spectrum, snow is the same reflectance as clouds, making it difficult to differentiate between the two. By incorporating the secondary band (the SWIR), the differences appear stronger than just the visible alone. Snow absorbs sunlight at a different frequency than clouds, at 1.6 microns. Values in the NDSI that are greater than 0.4 microns are typically indicators of snow presence. (Earth Observation System, 2019; Nijhawan, Garg and Thakur 2016)
NDSI is a significant method used in the landcover analysis process. It allows for a spectral discrimination between snow, soil, rock, and cloud cover. This is an extremely useful tool when mapping snow in rough terrain.
NDSI is determined using digital numbers of two TM bands using the following equation (Equation 1):
NDSI results in an image with a sharp boundary at the terminus of the glacier and surrounding moraine. It also allows for multitemporal comparison of the glacial tongues. The outside limits of glaciers are never located in topographic shadows, providing a great means to delineate glacial boundaries. (Silverio and Jaquet 2005)
Change Detection and Binary Change
Change detection analysis was inspired by Engeset et al., 2002 and Nijhawan et al., 2016 articles from different parts of the world, but the methods may be applied elsewhere. Change detection is done by taking the brightness values from one image from another ranging from 0-255. In my case, I wanted to see the difference between the near-infrared portion of the spectrum for each timestep image, as it was the first to give me usable results. A constant is applied to counteract any possible negative values, as a negative value still represents a change of some sort.
Equation 2 above is the formula used to create the change detection images. (Engeset, et al. 2002; Nijhawan, Garg and Thakur 2016)
Binary Change
To find areas of significant changes in the data, we use a conditional function to weed out the insignificant change data. The change threshold value for each image is done by taking the input image’s mean value and adding it to its standard deviation. Equation 3 is the conditional statement utilized to separate the image values.
Each image’s change threshold will define the amount of pixel values accepted to fit within the acceptable threshold to show change, or not fit within the acceptable threshold.
Results
NDSI
Figure 7 shows the results of the NDSI image analysis. The images show areas with lighter shading indicate more snow/ice and darker with less. As the image suggests, the southern border and upper notch of the glacier consistently contain high amounts of snow/ice.
Change Detection
The results of the change detection formula from equation 2 are noted in table 3. The change threshold values and image input names from table 3 were used in processing the results in the binary change detection formula from equation 3.
Binary Change Analysis and Quantification of Decline
The results of the binary change detection processed into an overlay map is shown in figure 8.
Values of 0 change are excluded from the map, as they were not significant based on their change threshold values.
In order to calculate how much change occurred in the areas we denoted with the ‘1,’ the pixels must be converted to real-world numbers. Table 4 shows the pixel values and their coding along with the date. Because the image resolution was 30 x 30 in the chart earlier, the ‘1 change’ pixels are multiplied by 900 to get the number of square meters that have changed over each timestep. Conversion to square meters to acres was intended for ease of reading through utilization of a conversion calculator. The results of these calculations are in table 5.
Discussion
My expectations of the project were that there would be a more noticeable change between the years 2000 to 2015 than 1985 to 2000. However, the results of my analysis show that the Gries glacier had experienced more change between 1985 and 2000 than from 2000 to 2015. This may be due to protocols and international actions to combat climate change in the 1990s causing a reduction in the amount of greenhouse gases in the atmosphere. These differences could have been a result of several factors, as described in the limitations section.
Limitations
Limitations to data and methods vary during different aspects of analysis. Remote sensing imagery qualities vary between sensors and may produce different ranges of brightness values. This makes comparison between different date and satellite projects difficult to distinguish. The imagery is also limited by date as to when the satellite was over the study area and the atmospheric conditions that the area is experiencing at the time. A larger range of dates or seasonality may have produced more complete results.
My results for the binary change analysis were based on an area of interest provided by a shapefile from the extent of the Gries circa 1992. This would have an influence on the pixels used to identify areas of change from the change detection as well as my change detection binary coding results. The change threshold depends on the amount of standard deviations specified by the study as well. In other projects, the standard deviation was multiplied by three. In this case, that result wouldn’t have existed on the histograms of the images. Therefore, the change threshold is more arbitrary and dependent on manipulation by the analyst.
Most of the mass balance and other measurements provided in this analysis was secondary rather than primary data. It would be worth pursuing through ground truthing and taking measurements for primary data. Another limitation regarding climate and record data was the fact that a large portion of the Gries-specific articles and data were in languages used in Switzerland. Use of a translator or transcriber would be immensely helpful and take up less of the time necessary to complete a project of this caliber. The use of international data is another issue that was faced. Only free to the public and English data was able to be used in this analysis.
Conclusion
The significance of glaciers comes from their “canary in the coalmine” behavior towards climate change; the glaciers’ importance in scientific endeavors through the data stored in atmospheric particles suspended in the melting ice. It is also important to note that the Gries glacier is one of thousands of glaciers each going through their own shifts and reactions to climate change variables. Glaciers like the Gries are a major part of valley life for much of Italy and Switzerland as well as other communities downstream of the Rhône and Rhine Rivers. Life as we know it will be impacted by a changing climate, whether we pay attention to it or not. (Levermann, et al. 2012)
Recommendations and Further Analysis
I’d like to work on finding further correlations between variables and including others that I had not considered/had access to for my project. It would be interesting to see how future predictions of climate change impacts on the Gries would change depending on various action plans. Climate action plans or protocols that are based on continuing life and business as usual, making moderate life changes, or making significant lifestyle changes. Application of different time scales would have a significant difference in analysis compared to this study. Also, possible collaborations with scientists from different disciplines and their insights on the issues relating to climate change could produce worthwhile results.
Bibliography
Earth Observation System. 2019. Ice Cover Monitoring. https://eos.com/environment-pollution-and-climate/ice-cover-monitoring/.
—. 2019. Snow and Cloud. https://eos.com/snow-and-cloud/.
Engeset, R. V., J. Kohler, K. Melvold, and B. Lundén. 2002. "Change detection and monitoring of glacier mass balance and facies using ERS SAR winter images over Svalbard." International Journal of Remote Sensing (Taylor & Francis Group) 23 (10): 2023-2050. Accessed April 3, 2019. doi:http://dx.doi.org/10.1080/01431160110075550.
Fischer, M., M. Huss, and M. Hoelzle. 2015. "Surface elevation and mass changes of all Swiss glaciers 1980-2010." The Cryosphere (Copernicus Publications) 9: 525-540. Accessed April 15, 2019. doi:http://dx.doi.org/10.5194/tc-9-525-2015.
Funk, Martin. 2007. Gries Glacier: From the Glacier Photograph Collection. National Snow and Ice Data Center, Boulder, Colorado.
Huss, M. 2012. "Extrapolating glacier mass balance to the mountain-range scale: the European Alps 1900-2100." The Cryosphere (Copernicus Publications) 6: 713-727. Accessed March 18, 2019. doi:http://dx.doi.org/10.5194/tc-6-713-2012.
Huss, M., A. Bauder, M. Funk, and R. Hock. 2008. "Determination of the seasonal mass balance of four Alpine glaciers since 1865." Journal of Geophysical Research 113: 1-11. Accessed May 21, 2019. doi:http://dx.doi.org/10.1029/2007JF000803.
Kääb, Andreas, Horst Machguth, and Frank Paul. 2005. "On the impact of glacier albedo under conditions of extreme glacier melt: the summer of 2003 in the Alps." EARSeL EProceedings. Zurich, Switzerland. 139-148.
Kappenberger, Giovanni. 2007. Gries Glacier: From the Glacier Photograph Collection. National Snow and Ice Data Center, Boulder, Colorado.
Levermann, Anders, Jonathan L. Bamber, Sybren Drijfhout, Andrey Ganopolski, Wilfried Haeberli, Neil R. P. Harris, Matthias Huss, et al. 2012. "Potential climatic transitions with profound impact on Europe: Review of the current state of six 'tipping elements of the climate system'." Climatic Change, February: 845-878. Accessed March 28, 2019. doi:http://dx.doi.org/10.1007/s10584-011-0126-5.
Naegeli, Kathrin, and Matthias Huss. 2017. "Sensitivity of mountain glacier mass balance to changes in bare-ice albedo." Annals of Glaciology 58 (75pt2): 119-129. Accessed April 15, 2019. doi:http://dx.doi.org/10.1017/aog.2017.25.
Nijhawan, Rahul, Pradeep Garg, and Praveen Thakur. 2016. "A comparison of classification techniques for glacier change detection using multispectral images." Perspectives in Science, September: 377-380. Accessed April 1, 2019. doi:http://dx.doi.org/10.1016/j.pisc.2016.04.080.
Paul, Frank, and Wilfried Haeberli. 2008. "Spatial variability of glacier elevation changes in the Swiss Alps obtained from two digital elevation models." Geophysical Research Letters 35 (21): 1-5. Accessed May 21, 2019. doi:http://dx.doi.org/10.1029/2008GL034718.
Paul, Frank, Horst Machguth, and Andreas Kääb. 2005. "On the impact of glacier albedo under conditions of extreme glacier melt: the summer of 2003 in the Alps." EARSeL eProceedings 4: 139-148. Accessed April 3, 2019. https://www.researchgate.net/publication/237240999_On_the_impact_of_glacier_albedo_under_conditions_of_extreme_glacier_melt_The_summer_of_2003_in_the_ALPS.
Rebetez, M. 1996. "Seasonal relationship between temperature, precipitation and snow cover ina mounainous region." Theoretical and Applied Climatology, July 17: 99-106. Accessed April 3, 2019. doi:https://dx.doi.org/10.1007/BF00865152.
Silverio, Walter, and Jean-Michel Jaquet. 2005. "Glacial cover mapping (1987-1996) of the Cordillera Blanca (Peru) using satellite imagery." Remote Sensing of Environment, 342-350. Accessed May 21, 2019. doi:http://dx.doi.org/10.1016/j.rse.2004.12.012.
Vincent, C., C. Fischer, C. Mayer, A. Bauder, S. P. Galos, M. Funk, E. Thibert, D. Six, L. Braun, and M. Huss. 2017. "Common climatic signal from glaciers in the European Alps over the last 50 years." Geophysical Research LEtters, February 26, 44 ed.: 1376-1383. Accessed May 28, 2019. doi:http://dx.doi.org/10.1002/2016GL072094.
World Glacier Monitoring Service. 2009. Global Glacier Changes: Facts and Figures. Copenhagen: UNEP.
Zemp, M., F. Paul, M. Hoelzle, and W. Haeberli. 2008. "Glacier fluctuations in the European Alps, 1850-2000: an overview and spatio-temporal analysis of available data." In Darkening Peaks: Glacier Retreat, Science, and Society, by B. Orlove and et al., 152-167. Berkeley, US: Zurich Open Repository and Archive, University of Zurich. Accessed March 31, 2019. doi:http://dx.doi.org/10.5167/uzh-9024.
Comments