SNOWDEPTH - Global snow depths from spaceborne remote sensing for permafrost, high-elevation precipitation, and climate reanalyses

SNOWDEPTH combines laser measurements from the ICESat-2 satellite with other satellite data, maps, climate reanalyses, elevation data, and statistical methods to map the amount of snow on the ground. Snow depth maps currently only exist for a few areas in the world, as there is no efficient method for measuring snow depth over larger areas, especially in remote regions. The amount of snow in winter is not only useful to know for skiing conditions, but also for the ecosystem, ground permafrost, and, importantly, it determines the amount of melt water in rivers during spring and summer. Mountain snow is a critical source of drinking water, hydro power, and irrigation, but is also a source of floods for large parts of the global population. Field measurements, such as from weather stations, primarily exist in easily accessible locations in wealthy countries. The SNOWDEPTH project provides currently unavailable, sought-after global information on snow depth. This data will be useful for many related disciplines in Norway and is particularly significant for less developed countries and areas where there is a sparse network of measurement stations. In the first part of the project, we developed methods to extract snow depth information from ICESat-2 data. The snow depth data from this satellite consists of profiles along straight lines, rather than maps of an entire area. The satellite takes measurements in the same area only about every three months and never at precisely the same location. We have found that these snow profiles can be as accurate as measurements taken with traditional methods and drones, even in challenging areas such as forests, where vegetation obstructs the snow on the ground. To obtain accurate snow data, a thorough understanding of measurement errors in all datasets is necessary. We employed both customised manual data processing for smaller areas and statistical methods based on machine learning to remove errors in larger areas. As snow depth profiles are not as useful as maps, the project has developed methods to create time series of snow depth maps based on the snow depth profiles, various geodata, and snow models. These methods are based on ensemble-based data assimilation techniques, similar to those used for climate reanalyses. We also utilize machine learning based on spatial statistics, or a combination of both methods. These approaches are relevant contributions to the research field and will continue to be further developed. So far, we have produced snow depth map time series for two smaller areas in the Alps and the Pyrenees, as well as monthly snow depth maps for the Hardangervidda region. Compared to traditional methods (without snow depth information), our models have demonstrated better estimation of the correct snow amount in a hydrological catchment, especially in the first half of winter. The use of spatial statistics and machine learning allowed us to recreate the distribution of snow with varying depths in the terrain in a highly realistic manner. In the second part of the project, we will improve and expand the methods and create snow depth maps for more locations around the world, including areas where no similar measurements currently exist. Part two also includes three applications where snow depth has significant potential to contribute to new knowledge, also in the light of climate change: 1) permafrost: Snow insulates the ground from the cold winter air, and accurate maps of snow depth are a key factor for modelling permafrost correctly; 2) climate reanalyses: as these are based on field observations, climate reanalyses do not represent the past weather and climate very well in areas with few measurements. The snow depth measurements of SNOWDEPTH could thus improve climate reanalyses; and 3) high-elevation precipitation: precipitation processes at higher elevations are poorly understood due to the lack of direct measurements in mountain areas. Satellite-based snow depth data can help to fill this knowledge gap. The SNOWDEPTH project collaborates with researchers in Norway, Switzerland, Spain, and the United States. We conduct field studies in Norway, Central Asia, and Svalbard and have led an initiative to bring together snow researchers in Norway to form a collaborative network.

The SNOWDEPTH project will, as the first in the world, directly measure snow depths globally at high spatial resolution from freely available ICESat-2 spaceborne laser altimetry data available since autumn 2018. To generate global monthly snow depth maps, including for mountainous and forested areas, we will combine the ICESat-2-derived snow depths with Sentinel snow cover/depth data in an ensemble-based data assimilation (DA) framework. This global snow depth data will fill a large data and knowledge gap within hydrology and cryosphere/climate sciences and is directly relevant for the three application cases within the project: permafrost, high-elevation precipitation and climate reanalysis. The project has two parts and is supported by field activities for ground reference. In phase 1, we will develop algorithms to derive snow depths at two complementary scales: A) local snow depths from ICESat-2 profiles that capture the high spatial variability in areas with small-scale topography, and B) global snow depth maps with monthly temporal resolution, using DA methods. In phase 2, we will use the derived snow depths within three application fields where they directly benefit to advance the state of the art: i) Permafrost: include snow depths in an existing model framework to greatly improve modelling of the ground thermal regime, both locally at targeted field sites and at global scale. The current lack of snow depth data is a key bottleneck for permafrost modelling. ii) High-elevation precipitation: analyse how snow depths vary across orographic barriers to increase understanding of high-altitude precipitation processes. These are currently largely unconstrained due to lack of measurements. iii) Climate reanalysis: verify and improve operational and climate reanalysis products through cross-comparison and improved process understanding. In data-sparse areas, reanalysis products are less accurate and largely model-driven given the lack of observations.