ClustSnow

Version: 4.0
Author: Joschka Geissler
Last modified: 25 January 2025

Background

This documentation introduces the required steps for applying the ClustSnow model. ClustSnow determines snow distribution patterns from spatially distributed, multitemporal snow depth maps.

ClustSnow is implemented in R (Version 4.1.0) and requires the following libraries to be installed:

Required R-Packages

Name	Version	Literature
raster	3.4-13	Hijmans (2021)
stringr	1.5.0	Wickham (2009)
caret	6.0-88	Kuhn (2008)
nixmass	1.0.2	Winkler et al. (2021)

---

The ClustSnow workflow, introduced by Geissler et al. (2023) and Geissler et al. (2024), derives daily snow depth (HS) and snow water equivalent (SWE) maps based on observations.

Key steps include:

1. Spatial clustering: Derive spatial clusters (areas with similar snow dynamics) using the getCluster() function.
2. Cluster ordering: Order clusters based on their mean snow depth using the orderCluster() function.
3. Synthetic HS time series: Generate daily synthetic snow depth time series for each cluster with the hs.synth() function.
4. SWE time series: Convert synthetic HS time series to SWE time series using the delta.swe model.
5. Mapping: Extrapolate synthetic time series into space to create spatiotemporally continuous SWE and HS maps using the createMap() function.

Important Notes:

• SWE maps are only derived if time series are continuous and snow depth time series start with 0 m.
• Ensure proper directory structure as required:

---

Required Input Data and Format

1. Time Series

• Name: hs_time_series.csv
• Separator: ;
• Decimal Delimiter: .
• Additional Note: Specify the date format in the R script.

2. Sensor Locations

• Name: hs_sensor_location.shp
• Description: Shapefile containing all sensor locations. Indicate the attribute name containing sensor IDs in the R script. Sensor IDs must match the headers in the time series file.

3. HS Maps

• Name: hs_raster_stack.tif
• Description: Raster stack containing spatial maps of snow depth (e.g., UAV-based LiDAR).

---

Additional Tools

CloudCompare_CoregisterLiDAR.R

• Purpose: Co-registration of LiDAR-derived point clouds using Cloud Compare's Command Line Mode.
• Description: Wrapper script for running the co-registration workflow in R.
• Notes:
  • Adapt directories, AOI, and filtering thresholds to individual datasets.
  • Refer to Geissler et al. (2023) for more information.

---

References

1. Geissler, J., Mazzotti, G., Rathmann, L., Webster, C., & Weiler, M. (2024). ClustSnow: Utilizing temporally persistent forest snow patterns under variable environmental conditions. DOI:10.22541/essoar.172222597.78203131/v1
2. Geissler, J., Rathmann, L., & Weiler, M. (2023). Combining Daily Sensor Observations and Spatial LiDAR Data for Mapping Snow Water Equivalent in a Sub‐Alpine Forest. Water Resources Research, 59(9), Article e2023WR034460. DOI:10.1029/2023WR034460
3. Hijmans, R. J. (2021). raster: Geographic Data Analysis and Modeling. CRAN
4. Kuhn, M. (2008). Building Predictive Models in R Using the caret Package. Journal of Statistical Software, 28(5). DOI:10.18637/jss.v028.i05
5. Wickham, H. (2009). stringr: CRAN Contributed Packages. DOI:10.32614/CRAN.package.stringr
6. Winkler, M., Schellander, H., & Gruber, S. (2021). Snow water equivalents exclusively from snow depths and their temporal changes: the Δsnow model. Hydrology and Earth System Sciences, 25(3), 1165–1187. DOI:10.5194/hess-25-1165-2021