Since starting our in-house remote sensing program in 2016, we have used UAVs (Unmanned Aerial Vehicles or “drones”) to collect LiDAR and photogrammetry data on dozens of different types of projects: road corridors, urban city blocks, trailheads, reservoirs, bridges, dams, pipelines, commercial developments—you name it, we’ve likely flown a drone over it! Unique project requests happen regularly given the diversity of our clients and partners, but after a conversation in early 2020 with NASA research scientist Dr. Carrie Vuyovich we knew we’d be embarking on something different. Over the next few months, we began working on our most unique UAV remote sensing project to date: LiDAR snow surface mapping for the NASA SnowEx program.

SnowEx and the Future of Snowpack Monitoring  

The SnowEx program seeks to develop remote sensing methods for measuring the Earth’s annual snowpack and its associated hydrology. Billions of people depend on snowmelt for drinking water, irrigation, wildland fire-fighting, power generation, and outdoor recreation. As annual snow patterns continue to be altered by our changing climate, we must prepare to address severe consequences to our water resources. However, we don’t currently have a good set of tools to measure snow water content (referred to as “Snow Water Equivalent” or SWE) across the globe. This severely limits our ability to monitor water as a precious resource, understand how it’s changing, and ultimately determine how we can mitigate consequences and adapt to a changing planet.

The MSU Central Agricultural Research Center (CARC) in Moccasin, Montana served as an experimental SnowEx site for our LiDAR snow surface mapping.

Remote sensing tools such as LiDAR, imagery, and Synthetic Aperture RADAR (SAR) sensors mounted on aircraft and satellites can be used to capture snow measurements, but their application for this purpose is still in its relative infancy. Since 2016, SnowEx has been conducting winter field campaigns to determine the best types of sensors, data collection practices, and analysis techniques for monitoring snow in different biomes. The technologies and knowledge developed by SnowEx will lay the groundwork for the future design of a satellite system dedicated to monitoring snow dynamics—and the associated water that snow provides—across the globe.

LiDAR Mapping Prairie Snowpack in Montana

For our part in the 2021 field campaign, DJ&A worked with Dr. Eric Sproles of Montana State University Department of Earth Sciences to collect UAV LiDAR data on a 1km2 study site near Moccasin, Montana. The site served to model a prairie biome, one of the most common snow-covered biomes on Earth. To collect detailed measurements of changing snow depths throughout the season, we began by establishing a survey-grade control network and then collected UAV LiDAR prior to the first snowfall to establish a baseline ground surface. DJ&A field crews then revisited the site seven times from January to March to collect additional LiDAR datasets, capturing changes in the snow surface due to snow accumulation, melting, and displacement due to wind.

Preparing to take off with our LiDAR UAV equipped with a specialized 905nm LiDAR system for snow surface mapping.

Accurately mapping snow surfaces requires specialized equipment and rigorous survey procedures. We routinely use an advanced LiDAR system (based on the Riegl VUX-1 LiDAR sensor) for standard topographic mapping, but because snow crystals absorb and scatter the laser light used in most LiDAR systems, we required a specialized LiDAR sensor for this project. To successfully collect snow surface data, we called upon our trusted partner, Phoenix LiDAR Systems for support. As leaders in aerial LiDAR system design, they were able to provide a specialized solution to fit the needs of this unique project: a 905nm infrared LiDAR sensor.

Our LiDAR UAV lifts off to begin collecting snow surface mapping data. The LiDAR system collects thousands of elevation points every second.

Data management is always a challenge on UAV projects. Collecting LiDAR and high-resolution image files for tens or hundreds of acres can easily generate hundreds of gigabytes of data. We use GNARBOX intelligent hard drives to perform data backup and automatic file corruption checks in the field. This has become an important part of our workflow to ensure we leave UAV project sites confident that our data is intact and backed up. You can learn more about how we used the GNARBOX 2.0 Drone Edition on the SnowEx project in GNARBOX’s case study.

The UAV LiDAR Advantage

The combination of a specialized snow-compatible LiDAR system, high precision survey control and truthing methods, and meticulous point cloud processing, yielded impressive results. Six of our eight LiDAR surfaces achieved vertical accuracies that met or exceeded USGS Quality Level 1 (<0.64ft / 19.6cm RMSEz @ 95% confidence level) and two of the surfaces achieved USGS Quality Level 0, the highest possible, with vertical accuracies of <0.32ft / 9.8cm at a 95% confidence level. By comparing LiDAR snow surfaces from different days, we were able to detect changes in the snow surface as small as just a few centimeters (one tenth of a foot). Changes were especially apparent in areas where snow drifts formed due to wind, as displayed in the image below.

Cross section of LiDAR surface points collected along a windbreak before and after snowfall. Red points show the ground surface prior to snowfall. Blue points display the snow surface and show a snowbank building up along the windbreak.

Additionally, all of the point clouds had point densities exceeding 130 points/square meter. These statistics indicate an extremely high level of accuracy and detail for aerial LiDAR. It would be nearly impossible — and exorbitantly expensive — to achieve this level of accuracy with a traditional LiDAR system mounted on an airplane or helicopter. These results showcase the advantages a UAV LiDAR system for high-accuracy, high-detail surface mapping.

Cross section of all LiDAR point clouds near a ditch. Each color represents the snow surface on a different day, with red representing the ground prior to snowfall. Over time, blowing snow accumulated to multiple feet deep in ditches like this one.

The final snow surface point clouds were submitted to the SnowEx community and will be made available to dozens of scientists for use in their research of snow cover dynamics in prairie biomes. More information on all of the 2021 SnowEx activities can be found on NASA Goddard’s Earth Science News Page. We’re proud to have been able to play a small part in the innovative and globally important research being conducted by the SnowEx community.