SnowView
A Satellite Data and Model Driven Decision Support Tool for monitoring snowpack, precipitation, and streamflow.
Introduction
- Snowmelt from mountain forests is critical for water supply in Arizona and across the western US.
- Seasonal streamflow predictions often use observations that are not representative of surrounding areas.
- Long-term gridded snow data needed to infer snow condition across landscape to improve seasonal streamflow predictions.
The main watersheds that feed water into Phoenix include the Verde, Salt and Little Colorado.
UA/SWANN Data
- Based on the University of Arizona (UA) daily 4 km Snow Water Equivalent (SWE) and snow depth dataset.
- Daily data starting in 1981.
- Includes interpolated SWE estimates from ground stations (Broxton et al., 2016); physically based snow density model (Dawson et al., 2017).
- Data available from National Snow and Ice Data Center (NSIDC ).
Top: SNOTEL and COOP Stations where snow data is used in the creation of the UA SWE data set. Bottom: Average maximum SWE based on this data.
- UA/SWANN model: used to provide snow monitoring for Salt River Project (SRP).
- UA data downscaled to < 1 km using machine learning of SWE response to various physiographic indicators.
- Trained and evaluated with snow survey data from Arizona (Broxton et al., 2019) and elsewhere in the western US (Harpold et al., 2014).
- Generated in near-real time.
Top: Ground surface created from Drone acquried image data. Bottom: Researchers from the University of Arizona performing site based snow measurements.
Dataset Evaluation
- UA/SWANN data are compared to lidar coverages from the western US.
- At most sites, captures the spatial pattern of snow depth at most sites.
- Better simulation of snow depth than SNODAS at some sites; goes back much further in time.
Comparsion of SNODAS and UA-SWANN measurements for various sites across the southwest United States.
Comparison of various SnowDepth measurements including SNODAS, SWAAN and ASO. ASO - NASA Airborne Snow Observatory (Painter et al., 2016)
Streamflow Forecasting
- SnowView datasets improve seasonal streamflow forecasts (e.g. to help SRP plan reservoir operations).
When based on these gridded snow data (red line), they are better than the current state of the art of using station snow data (blue line). Credit: Bo Svoma, SRP
- Forecasts based on UA/SWANN SWE data, climatic indices, long range seasonal forecasts, and other land antecedent conditions.
- Machine learning to make predictions of remaining streamflow for every calendar day.
- Models trained with data from 1982, run for every year since then.
SnowView Decision Support Tool
SnowView is used to explore snow metrics both geographically and graphically over time.
- Helps Salt River Project (SRP) to determine how current year’s water supply might compare with previous years.
SnowView contains an array of data sets to explore the varius aspects of snow at the region scale.
- SWE Data: UA / SWANN, SNODAS data
- Snowcover Data: MODIS imagery, IMS data
- Precipitation Data: PRISM data
- Point Data: SNOTEL SWE and Precipitation data, USGS Streamflow data
- Expanded to include similar information and functionality for additional watersheds across the US.
- Map visualization across the US as well as timeseries data for USGS HUC2,4,6,8 watersheds, SNOTEL stations, and USGS stream gauges.
Summary
- Gridded SWE data offer advantages over point SWE observations for water supply modelling because they can estimate SWE conditions across the landscape.
- The UA/SWANN SWE data provide near real-time SWE estimates that can be used for water supply modelling.
- They improve SRP seasonal streamflow forecasts over using only SNOTEL data.
- Currently, we are using UA/SWANN SWE data to develop daily updating seasonal streamflow forecasts for SRP4 km SWE dataset (1981-2017) can be downloaded from NSIDC.
- Higher Resolution data can be visualized using our SnowView interface.
References
- Broxton, P., X. Zeng, and N. Dawson. 2019: Daily 4 km Gridded SWE and Snow Depth from Assimilated In-Situ and Modeled Data over the Conterminous US, Version 1. Boulder, Colorado USA. NASA National Snow and Ice Data Center Distributed Active Archive Center. doi:10.5067/0GGPB220EX6A.
- Broxton, P. D., W. van Leeuwen, J. Biederman, 2019: Improving Snow Water Equivalent Maps with Machine Learning of Snow Survey and Lidar Measurements, Water Resources Research, doi:10.1029/2018WR024146.
- Dawson, N., P. D. Broxton, X. Zeng, 2017: A new snow density parameterization for land data initialization. J. Hydrometeor., 18, 197-207, doi:10.1175/JHM-D-16-0166.1.
- Broxton, P. D., X. Zeng, N. Dawson, 2016: Why Do Global Reanalyses and Land Data Assimilation Products Underestimate Snow Water Equivalent? J. Hydrometeor., 17, 2743–2761, doi:10.1175/JHM-D-16-0056.1.
- Painter, T. H., Berisford, D. F., Boardman, J. W., Bormann, K. J., Deems, J. S., Gehrke, et al., 2016: The Airborne Snow Observatory: Fusion of scanning lidar, imaging spectrometer, and physically-based modeling for mapping snow water equivalent and snow albedo. Remote Sensing of Environment, 184, 139-152.
- Harpold, A. A., Q. Guo, N. Molotch, P. D. Brooks, R. Bales, J. C. Fernandez‐Diaz, K. N. Musselman et al., 2014: LiDAR‐derived snowpack data sets from mixed conifer forests across the Western United States. Water Resources Research, 50(3), 2749-2755.