Decision tree-based detection of blowing snow events in the European Alps

Abstract

Abstract. Blowing snow processes are crucial in shaping the strongly heterogeneous spatiotemporal distribution of snow, and in regulating subsequent snowpack evolution in mountainous terrain. Although empirical formulae and a constant threshold wind speed have been widely used to estimate the occurrence of blowing snow in regions with sparse observations, the scarcity of in-situ observations in mountainous regions contrasts with the demands of models for reliable observations at high spatiotemporal resolution. Therefore, these methods struggle to accurately capture the high local variability of blowing snow. This study investigated the potential capability of the decision tree model (DTM) to detect blowing snow in the European Alps. The DTMs were constructed based on routine meteorological observations (mean wind speed, maximum wind speed, air temperature and relative humidity). Twenty repetitions of random sub-sampling validation test with an optimal size ratio (0.8) between the training and validation subset were applied to train and assess the DTMs. Results show that the maximum wind speed contributes most to the classification accuracy, and the inclusion of more predictor variables improves the overall accuracy. However, the spatiotemporal transferability of the DTM might be limited if divergent distributions exist between stations. Although both the site-specific DTMs and site-independent DTM show strong performance for accurately detecting blowing snow, specific assessment indicators varied between stations and surface conditions. Events for which blowing snow and snowfall occurred simultaneously were detected the most reliably. Although models failed to fully reproduce the high frequency of local blowing snow events, they have been demonstrated a promising approach requiring limited meteorological variables and have the potential to scale to multiple stations across different regions.

A Framework for Estimating the Probability Distribution of Event Runoff Coefficient in Ungauged Catchments

“…Principal components analysis (PCA; Krzanowski, 2000), information gain (IG; Ssegane et al., 2012a), and recursive feature elimination (RFE; Demarchi et al., 2020) are common and practical variable selection methods. After screening out the less informative catchment attributes, machine learning methods (such as a classification tree (Ragettli et al., 2017; Singh et al., 2014; Xie et al., 2021), random forest (Addor et al., 2018; Stein et al., 2021)), or a regression model (Bloomfield et al., 2021; Tarasova, Basso, Poncelet, et al., 2018) could be adopted to assess which catchment attributes are the influential factors for indicating the cluster categories with similar ERC characteristics.…”

Snow depth variability across the Qinghai Plateau and its influencing factors during 1980–2018

“…Local terrain characteristics such as slope, orientation and shadowing influence snow accumulation (Jain et al ., 2008). Wind can shape both the spatial heterogeneity of snowfall and erosion, transport and deposition of surface snow via blowing snow (Xie et al ., 2021). However, few studies examined and quantified the effects of wind on snow depth variation on a large scale.…”