HS6.4

Satellite observations are the only means for timely and complete observations of the global snow cover. A range of different satellite snow products is available, the performance of which is of vital interest for the global user community. We provide an overview on goals and activities of the SnowPEx+ initiative, dedicated to the intercomparison of northern hemispheric and global satellite snow products, derived from data of long-term operational as well as recently launched satellites. SnowPEx+ is the continuation of SnowPEx (2014-2017), carried out as an international collaborative effort under the umbrella of Global Cryosphere Watch / WMO and funded by ESA.

SnowPEx+ focuses on two parameters of the seasonal snowpack, the snow extent (SE) from medium resolution optical satellite data (Sentinel-3, VIIRS, MODIS, AVHRR, etc.) and the snow water equivalent (SWE) from passive microwave satellite data. Overall, 15 hemispheric and global SE products (binary and fractional SE) and two SWE products are participating in the experiment. For intercomparison, daily SE products are transformed to a common map projection and standardized SnowPEx protocols are applied, elaborated by the international snow product community. The SE product evaluation applies statistical measures for quantifying the agreement between the various products, including the analysis of spatial patterns. Validation of SE products uses as benchmark high resolution snow maps from about 150 globally distributed Landsat scenes acquired in different climate zones, under different solar illumination conditions and over various land cover types. This snow reference data set, based on various retrieval algorithms, is generated and evaluated by the SnowPEx+ High Resolution Snow Products Focus Group. In-situ snow data from several organisations in Europe, North America and Asia are also used for validating the satellite SE and SWE products. SWE products are also inter-compared with gridded snow products from land surface models driven by atmospheric reanalysis data. In addition, the multi-year trends of the various SE and SWE products are evaluated. We provide an overview on the snow products, discuss the validation and intercomparison protocols, and report on preliminary results from the intercomparison and validation of various snow products.

How to cite: Nagler, T. and the SnowPEx Team: SnowPEx+: The International Snow Products Intercomparison and Evaluation Excercise 2015-2020, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-7319, https://doi.org/10.5194/egusphere-egu22-7319, 2022.

Precise snow cover estimations are relevant for many fields of research applications, such as for hydrological and meteorological modelling. Furthermore, snow plays an important role in hydropower management and flood prediction. Snow cover monitoring from satellite imagery has received increasing attention over the past decades. Nowadays, improvements in estimation methodologies and better availability of augmented satellite imagery provide an excellent basis for reliable estimations of fractional snow cover.

In this work we exploit the available spectral information of the Sentinel-2 MSI and Sentinel-3 OLCI for automatic estimation of fractional snow coverage. This is achieved through linear spectral unmixing with local endmembers. Similar implementation of methods that employ spectral unmixing use a reflectance model or a spectral library with pre-selected endmembers. Our approach selects the spectral endmembers from the data itself and applies them depending on the distance from the query point assuming spectral similarities of ground reflectance nearby. Endmembers are found through a pre-classification step based on conservative thresholds in combination with a similarity measure. The linear unmixing problem is solved several times for each query point using different combinations of endmembers detected in the vicinity of the query point; accounting for different illumination conditions and shaded areas. Finally, a careful selection of accurate fractional snow cover estimations is performed. This approach is globally applicable, adjusts to the local environment and illumination conditions and avoids costly endmember modelling or the provision of an external spectral library. The method was tested in different regions in the world using different satellite data including Sentinel-2, Landsat and Sentinel-3 OLCI and were inter-compared with snow information from other sources. In the presentation we will present the method, and examples of fractional snow cover maps. The performance of the method will be shown in comparison with other data and the limitations and capabilities will be discussed.

How to cite: Keuris, L., Nagler, T., Mölg, N., and Scheiblauer, S.: Fractional snow cover estimation through linear spectral unmixing of Sentinel-2 and Sentinel-3 optical satellite data using local endmembers, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-9932, https://doi.org/10.5194/egusphere-egu22-9932, 2022.

Daily snow cover is an essential parameter in hydrology, climate, and environmental studies. Although remote sensing images provide valuable information on snow, they are restricted by clouds, clouds’ shadows, and temporal and spatial coverage. This study synthesizes daily snow cover maps based on climate and near clear sky Sentinel-2 and Landsat images. The motivation of this study is that snow patterns are repeatable between years with similar meteorological characteristics. Accordingly, a distance metric based on climate information is computed, including temperature and precipitation (1km resolution) as well as auxiliary data such as daily MODIS snow cover. This distance quantifies the mismatch between the days when clear sky Landsat or Sentinel-2 data is available, learning days, and days when there is no clear sky satellite data or test days. The proposed methodology is applied on a subset of the Alpine belt called the Western Swiss Alps and on the Jonschwil sub-basin, both located in Switzerland. We have synthesized daily snow cover maps for each of our regions of interest for 20 years since 2000.

To evaluate synthesized snow cover maps, we use leave-one-out cross-validation, comparison with a random selection process, and a degree-day snow model. The leave-one-out assessment shows a good agreement between the actual Landsat and the synthesized one. The synthesized snow cover maps also show that the proposed method output agrees with physical concepts as the physical features have been used along with satellite data in the proposed model. Considering physical features in synthesizing Landsat images is an innovation that allows us to use the methodology to synthesize images for the pre-satellite period. Moreover, random selection assessment shows that considering a metric based on climate and auxiliary data can synthesize snow cover as repeatable patterns depending on meteorological data.

How to cite: Zakeri, F. and Mariethoz, G.: Synthesizing Daily Snow Cover Maps Using Satellite Images and Climate Information, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-11519, https://doi.org/10.5194/egusphere-egu22-11519, 2022.

Snow is an important hydrological component in Central Asia. The snowmelt contributes to about 50 % of total water formation in the region, depending on geographic conditions. Many hydro-meteorological phenomena such as floods or drought conditions can be triggered by snowmelt amounts in Central Asia. The amount of snow accumulation in the mountains of Tian-Shan and Pamir also defines the availability of water for summer months to be used for agricultural production or re-filling of reservoirs for energy production in the winter period. Thus, it is of high importance to better understand the seasonal variation of snow and if the over the global average climate warming in the region is affecting the processes related to snow accumulation and melt.

In this study, we analyze 22 years of daily Moderate Resolution Imaging Radiometer (MODIS) snow cover data that was processed using the MODSNOW-Tool, including cloud elimination. Additionally, observed snow depth data from meteorological stations were used to estimate trends related to snow cover change. We used several parameters such as snow cover duration, snow depth, snow cover extent, and snowline elevation to analyze changes.  We conducted this analysis in 18 river basins across the Central Asian domain with each river basin having different geographic conditions and the results show varying tendencies. In many river basins, a clear decrease of snow cover was found to be significant, whereas in some river basins also increase in the snow cover extent in particular months could be identified. We attributed the changes related to snow cover to available historical temperature and precipitation records from meteorological stations to better understand the driving forces. The results of this study indicate seasonal snow cover variations but also potential water shortages in particular months as well as water abundance in months where water demand is not high in Central Asia. 

How to cite: Gafurov, A., Kalashnikova, O., Niyazov, D., Mamaraimov, A., Gafurov, A., and Adkhamov, U.: Climate change-driven seasonal snow cover variations in Central Asia, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-12576, https://doi.org/10.5194/egusphere-egu22-12576, 2022.

Observations worldwide identify snow cover persistence together with snowfall occurrence as the most affected variables by global warming. In particular, Mediterranean mountain areas are pointed as climate warming hotspots. The characteristic snow-patched distribution shown over these areas, which result in different accumulation-ablation cycles during the cold season, usually makes spatial resolution the limiting factor for its correct representation. Remote sensing is the only feasible data source for distributed quantification of snow in mountain regions on medium to large scales, due among other to the limited access to these areas together with the lack of dense ground monitoring stations for snow variables. Among the numerous remote sensing sources, the Landsat constellation is those that better fit both basic requirements for studying snow over these areas, to cover a long period with observation and to have an high spatial resolution. However, the traditional classification algorithms for snow detection are usually based on normalized indexes that  provide a binary classification as snow and no-snow pixels throughout the study area; this simple classification may result in large error in heterogeneous and transitional areas within the snow-dominated domain. Alternatively, the spectral mixture analysis (SMA) approach provides a fraction of snow cover within each pixel and thus, constitutes a step forward to characterize heterogeneous and patchy snow areas in semiarid regions. 

This work analysed the role of mixed pixels, defined as pixels made up of different types of surfaces, in snow cover distribution over Mediterranean mountains. Sierra Nevada Mountain Range in southern Spain has been chosen as representative of a Mediterranean mountain area, which is characterized by strong altitudinal gradients with marked differences between the south (directly affected to the sea) and the north faces are found in the area. The fractional snow cover maps, at 30×30 m and 16 days spatial and temporal resolution respectively, derived from SMA of Landsat TM and ETM+ validated using as high resolution terrestrial photography (Pimentel et al., 2017) has been used for mixed pixel analysis. On the one hand, the results show the importance of mixed pixels, which can constitute more than 50% of the total pixels in some areas of the mountainous range and season of the year. On the other hand, the analysis carried out has allowed the identification of areas more prone to allocate this type of pixels, linking that fact to climatic drivers. 

This work has been funded by project MONADA - "Hydrometeorological trends in mountainous protected areas in Andalusia: examples of co-development of climatic services for strategies of adaptation to climatic change", with the economic collaboration of the European Funding for Rural Development (FEDER) and the Andalusian Ministry of Economic Transformation, Industry, Knowledge and Universities. R+D+i project 2020.

How to cite: Pimentel, R., Aparicio, J., Torralbo, P., Contreras, E., Moreno-Pérez, F., Aguilar, C., and Polo, M. J.: Quantifying the role of mixed pixels in snow cover distribution in semiarid regions: A study case in Sierra Mountain (Spain), EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-12801, https://doi.org/10.5194/egusphere-egu22-12801, 2022.

The inter and intra-annual dynamics of seasonal snow are of key interest in the tourism-based economies of many Alpine regions as well as for millions of people in the adjacent European lowlands when it comes to freshwater supply and electricity generation. However, accurate snow observations over long periods of time and at large spatial scales are especially challenging in inaccessible mountainous areas. This can be overcome by using data from Earth Observation satellites, which have been constantly monitoring the Earth’s surface for almost 40 years. On a catchment basis, we derive the Snow Line Elevation (SLE) from Landsat data for the entire Alpine region and model the spatio-temporal dynamics in monthly time-series ranging from 1984 to today. Based on the historical observations we model future SLE dynamics comparing different uni-variate and multi-variate approaches and assess them for their ability to generate multi-year forecasts from EO-derived time series data. These forecasts can enable local and regional stakeholders to adapt to a potentially changing snow regime under climate change.

How to cite: Köhler, J., Dietz, A., and Kuenzer, C.: Modeling future Snow Line Elevation dynamics in the Alps based on long remote sensing time-series, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-993, https://doi.org/10.5194/egusphere-egu22-993, 2022.

Abstract.

Snow depths are traditionally determined by point measurements at automatic weather stations or field observations, which cannot capture the complexity of snow depth distribution in alpine terrain. Therefore, remote sensing techniques have become key tools for spatially continuous snow depth mapping. Only satellites, airborne laser scanners (ALS) or photogrammetry from piloted aircrafts are capable of covering large regions of more than 100 km². However, the accuracy of satellite data does not match/achieve the requirements for exact snow depth mapping yet. In comparison to ALS, photogrammetric methods are considerably more economic, but have the disadvantages of light and weather dependence as well as the lacking ability to penetrate high vegetation. Nevertheless, previous studies of photogrammetric snow depth mapping on a large scale have already proven the accurate implementation, but those studies were either limited to only one recording or characterized by a spatial resolution of around 2 m. These properties limit the comparison of snow depth distribution and the analysis of small-scale features.

In our study we apply airborne imagery from the current state-of-the-art survey camera Vexcel Ultracam acquired during the annual peak of winter for the five years from 2017 to 2021 in an area of approximately 300 km² around Davos, Switzerland. This enables the calculation of outstandingly improved annual snow depth maps. The high spatial resolution of the snow depth maps (0.5 m) in combination with the high-resolution orthophoto (0.25 m) enables the identification of small-scale snow depth features. Additionally, the development of masks for high-vegetated and settled areas in combination with the high accuracy of the unmasked snow depth values (root mean square error of around 0.15 m) represents a significant step forward for reliable snow depth mapping of large alpine regions with photogrammetric methods. Our study focuses on the consistent workflow used for processing the snow depth maps, demonstrates the special characteristics of the snow depth distribution and presents the potential for investigations and applications based on this unique snow depth time-series.

How to cite: Bührle, L., Marty, M., Eberhard, L., Stoffel, A., and Bühler, Y.: Snow depth mapping over large, high-alpine regions by airplane photogrammetry, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-9744, https://doi.org/10.5194/egusphere-egu22-9744, 2022.

Rationale: After the Tropical Rainfall Measuring Mission (TRMM) was successfully launched by NASA and JAXA in 1997, NASA released the first GPM-era global precipitation product (IMERG) in April 2014, aiming to obtain precipitation data with ultra-fine temporal and spatial resolution around the world. Examining the key precipitation data in different climatic areas influenced by the monsoon can effectively help users and algorithm developers maximise the accuracy and characteristics of new satellite remote sensing products. Objective: To this end, this study used the upper and middle Lancang River basin (UMLRB), a transnational river with complex climatic conditions, as the research area to explore the applicability and precipitation distribution of IMERG and TRMM, and evaluate their accuracy. Methods: In this study, various performance indexes were used to comprehensively evaluate the retrieval accuracy of IMERG and TRMM remote sensing precipitation data in UMLRB; these indexes can be divided into two categories according to the evaluation objectives. One type of indexes mainly evaluates the amplitude consistency of precipitation, and the other type of indexes is mainly used to evaluate the occurrence consistency of precipitation. Results: The results indicated that: (1) The temporal distribution of precipitation in different climatic regions was correctly detected by IMERG and TRMM in the UMLRB, and the dry and wet seasons in the climate transition zone were distinct. (2) IMERG and TRMM tended to overestimate moderate rain (1.0-20 mm/d) while underestimating heavy rain (20-50mm/d) and extreme precipitation (> 50mm/d). (3) In terms of the amplitude consistency of precipitation, the detection results of IMERG in the alpine climate zone were not completely consistent with those of TRMM, while those in the climate transition zone were consistent with TRMM. (4) The stronger the precipitation intensity, the worse the accuracy of IMERG and TRMM, especially between heavy rain (20-50 mm/d) and extreme precipitation (> 50mm/d). (5) The IMERG, which had greater application potential in complex climatic conditions, had higher accuracy than TRMM.” Conclusions/Recommendations: Therefore, before using remote sensing precipitation data to study watershed hydrometeorology in monsoon-affected areas, their seasonal distribution, precipitation intensity, and the type of remote sensing data should be carefully considered to verify their accuracy.

How to cite: Lu, C., Fang, G., Ye, J., and Yang, Z.: Accuracy assessment of IMERG and TRMM remote sensing precipitation data under the influence of monsoon over the upper and middle Lancang River basin, China, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-2078, https://doi.org/10.5194/egusphere-egu22-2078, 2022.