Monitoring of the snow cover with optical remote sensing

Snow is a crucial environmental factor that reacts dynamically to climate change, especially in mountainous areas. Earth Observation (EO) data tracks snow cover at various spatial resolutions. Long-term remote sensing operations can create time series over several decades and identify long-term patterns.

A recently published study in remote sensing investigates the potential of remote sensing time series to predict snow line height in the European Alps. Researchers created time series of snowline elevation using data from the entire Landsat archive in 43 Alpine catchments.

Impact of snowpack dynamics on climate change and the environment

Snow is of vital environmental, social and economic importance in many regions of the world. Due to its high albedo, the snow cover directly influences climate change. Snow also has a significant impact on habitats and safeguards biodiversity.

Snow cover length is an effective indicator of species distribution, while snowmelt timing affects crop phenology and production. Habitat changes caused by snow cover dynamics caused by climate change increase the ambient temperature.

Importance of monitoring and estimating snow cover

For planners, tourism organizations and other stakeholders in the European Alps, it is crucial to have accurate predictions about the future dynamics of snow cover and snow line height. The high-altitude snowpack is a significant source of freshwater, enabling agriculture and energy production. Snow forms the basis for the tourism-based economy of several locations in Austria, Germany, France, Italy and Switzerland, particularly in the European Alps.

Alpine ski tourism in winter can be severely affected by decreasing snow cover duration and a progressively lower snow line height. Collecting environmental data over extended time scales and wide spatial scales depends heavily on space-based Earth Observation (EO).

Current techniques and products for monitoring snowpack dynamics

The formation of land surface data products from long RS time series is becoming more accessible with recent breakthroughs in remote sensing. The Global Snow Pack (GSP), a MODIS-based snow cover monitoring product, enables snow cover studies with a resolution of 500 m. Based on data from the Advanced High-Resolution Radiometer (AVHRR), even longer time series of snow cover dynamics have been created. The snow-related climatological and phenological investigation is particularly well suited for both sensors due to their extremely high temporal resolution.

Synthetic Aperture Radar (SAR) is a particularly promising technique for snow cover monitoring because its detection capabilities are independent of cloud cover and lighting conditions. SAR data are excellent for constructing gapless time series. Its ability to detect wet snow makes it easier to analyze the onset of seasonal snowmelt.

Landsat is another product for monitoring snow cover and facilitating the creation of optical time series. Although the data generated by Landsat has a lower temporal resolution (16 days) than MODIS or AVHRR, the far greater spatial resolution allows for thorough snow mapping even in complicated topography of high mountains.

Development of long RS time series to study snowpack dynamics

Most current projections of future snow cover are made using regional climate models (RCM) or general circulation models (GCM) with relatively coarse geographic resolution. Charcoal burner et al. developed the long RS time series to study historical snowpack dynamics and trends. Long RS time series are easier to construct when constructed from ARD due to reduced complexity, and they can provide a general indication of future dynamics at different spatial scales.

study results

Snow cover in mountainous areas is a metric heavily influenced by climate change. In this study, Koehler et al. created the basis for a snow line elevation prediction system based on long time series of EO data. Researchers created the first-ever Alpine-wide snow line elevation data set by generating monthly snow line elevation time series for each of the 43 Alpine river basins from EO data from 1985 to 2021.

The capacity of seven prediction algorithms to model and predict future snowline elevation based only on historical observations was evaluated using this dataset as input. Researchers projected future SLE time series through 2029 using a technique that integrates the best performing forecasts in each catchment. The expected movement in mean snowline elevation level maintained the long-term trend sign in the vast majority of the catchments.

Random Forest Telescope (0.76, 268 m) and seasonal ARIMA gave the best results with a Nash-Sutcliffe efficiency (NSE) of 0.79 and a mean absolute error (MAE) of 258 m. Significant gains in monitoring the Proposed snowline height can be predicted in the future by integrating external predictor factors into a multivariate modeling strategy.

Relation

Koehler J, Bauer A, Dietz AJ, & Kuenzer C (2022). Towards predicting future snow cover dynamics in the European Alps – the potential of long optical remote sensing time series. remote sensing, 14(18), 4461. https://www.mdpi.com/2072-4292/14/18/4461/htm

Disclaimer: The views expressed herein are those of the author, expressed in his own capacity and do not necessarily represent the views of AZoM.com Limited T/A AZoNetwork, the owner and operator of this site. This disclaimer is part of the terms of use of this website.