Forecasting permafrost thaw from aerial lidar requires projecting 3D point cloud features onto 2D prediction grids, yet naive aggregation methods destroy the vertical structure critical in forest environments where ground, understory, and canopy carry distinct information about subsurface conditions. We propose a projection decoder with learned height embeddings that enable height-dependent feature transformations, allowing the network to differentiate ground-level signals from canopy returns. Combined with stratified sampling that ensures all forest strata remain represented, our approach preserves the vertical information critical for predicting subsurface conditions. Our approach pairs this decoder with a Point Transformer V3 encoder to predict dense thaw depth maps from drone-collected lidar over boreal forest in interior Alaska. Experiments demonstrate that z-stratified projection outperforms standard averaging-based methods, particularly in areas with complex vertical vegetation structure. Our method enables scalable, high-resolution monitoring of permafrost degradation from readily deployable UAV platforms.

Mapping of magnetic anomalies in volcanic areas is a valuable tool to better understand lava flow geometries and dynamics. The high magnetization of basaltic lava allows us to easily identify buried lava flows, providing constraints on the total volume of old erupted material and the flow geometry, while magnetic mapping of volcanic intrusions in country rock enables us to model feeder dike geometry. The low magnetic signals within recent lava flows can identify areas that are still above the Curie temperature, constraining the dynamic of recent flow, and negative anomalies above old empty lava tubes can allow us to identify these hidden conduits. Traditionally, magnetic anomaly mapping for small regions is performed by walking through the survey area and for larger regions using crewed aircraft. Walking is often daunting, labor-intensive, and potentially dangerous. On the other hand, crewed aircraft are normally expensive and require a significant logistical organization before the survey. Uncrewed aerial vehicles (UAVs) have the potential to bridge the gap and collect a significant amount of high-resolution data in a relatively short time, possibly in areas not easily accessible. UAVs also provide the opportunity to collect data at multiple altitudes, providing a full gradient of the measured field. Here we present the results from both old and recent lava flows. Little Cones in Nevada (USA) consists of two visible cones erupted ~0.8Ma in the vicinity of the proposed nuclear repository of Yucca Mountain, Nevada. The lava flows from the two cones are partially buried by alluvium and not visible above ground. Our UAV data collection and data inversion allowed us to map the full extent of the lava flow and estimate the total volume of effused material. Our surveys of Hell’s Half Acre (~3000 BCE) lava flow, and Kings’ Bowl (~300 BCE) flow in Idaho (USA) are examples of the use of magnetic anomalies to identify lava tubes, feeding dikes, and flow morphology. Our 2022 survey of the 2018 Pacaya (Guatemala) lava flow is an example of a hard-to-access flow in which we can identify the warmer core that has not yet cooled below the Curie temperature. A survey of the 2018 Lower East Rift Zone eruption of Kīlauea in Hawaiʻi (USA) conducted in 2022 is another example of a negative anomaly possibly associated with areas that are still hot, or lava tubes. In all the surveys we collected a few hundred line kms of data in a few days using two 200 Hz triaxial fluxgate magnetometers mounted on a medium-lift drone. UAV magnetic surveys in volcanic regions are a powerful tool for understanding old and recent volcanic processes.

Monitoring coastal wetlands, particularly mangroves, is increasingly important as the impacts of climate change increase. As sea levels rise and temperature increase, vegetation communities traditionally associated with tropical and sub-tropical coastlines will migrate northward and also inland, along waterways. The transition from coastal marshes and subshrubs to woody mangroves is a fundamental change to coastal community structure and species composition, requiring monitoring. However, this transition is likely to be episodic, complicating monitoring efforts, as mangrove advances are countered by dieback resulting from increasingly impactful storms. Coastal habitat monitoring has traditionally been done through satellite and ground-based surveys. This project investigates the use of UAV lidar and multispectral photogrammetry which can be obtained routinely at higher resolution than satellite derived data, and cheaper and faster than ground-surveys. Using UAV-based methods we monitor and classify coastal habitats, including mangroves, using simple machine learning methods. Between 2020 and 2022 we investigated the use of remote sensing to monitor a multiple use Florida coastal ecosystem. Using UAV lidar we mapped vegetation communities and detected sections of significant canopy loss. Ground truthing verified the occurrence of recent canopy loss at the scale of individual snag remnants of woody mangrove associates, i.e., buttonwood trees (Conocarpus erectus). Using UAV lidar and multispectral photogrammetry data as inputs into a random forest model, we created several models of habitat classification. Training inputs included 2000-pixel and 5000-pixel data subsets. Initial results were resampled to match the size of tree crown in the field area creating four classification schemes. All classifications were validated using standard metrics. Mangrove habitat identification using the resampled 2000-pixel model has 85% producer’s accuracy and 80% user’s accuracy. UAV surveys combined with machine-learning streamline coastal habitat monitoring and facilitate repeat surveys to assess the effects of climate change.

Improvements in miniaturization and affordability of lidar technology, mainly due to innovation in self-driving cars, means that UAV lidar is now an accessible option for geoscience research. We present applications in which UAV lidar contributes to data collection in ways that would otherwise not be possible in the time frame, budget, and/or with the resolution required. Volcanoes: Lava flow surface texture can provide information on lava flow dynamics and emplacement. The transition between pahoehoe and a’a flow textures can indicate changes in flow rates and flow thickness, and the morphology of ripples in ropey pahoehoe flows can indicate flow direction. Hell’s Half Acre, Idaho, USA, is a basaltic lava flow that was erupted ~5000 y.a. Analysis of UAV lidar data at this lava field shows lava flow surface texture in sufficient resolution to define cm-scale pahoehoe ripples. In addition, larger scale lava features such as channels and inflation/deflation ridges can be mapped which allows us to understand the dynamics of the lava flow emplacement. Vegetation: UAV Lidar can be useful for analysis of vegetation canopy, both in stripping canopy (lidar last return) and in using it for tree height (lidar first return). By combining UAV lidar with other airborne data, e.g. multispectral imaging, we can identify and map tree species at Ft de Soto Park, in Florida, USA. Permafrost: Permafrost thermokarst features can develop rapidly and climate change will cause an increase in these rapid thaw events. With UAV lidar we can strip the vegetation to reveal the underlying ground surface which can then be used to assess and model permafrost processes. UAV surveys are quick and relatively inexpensive (as compared to crewed aviation) and data can be collected in response to a thaw event. We present data from Alaska, USA, at known sites of rapid thermokarst thaw. UAV lidar, both as a stand-alone dataset, and when integrated with other data streams e.g. multispectral and visible imagery, can provide high-resolution data (both spatial and temporal) on a platform that is relatively low-cost and logistically straightforward to deploy.

Barrier island beaches provide important protection, but human development, loss of natural cover, hurricanes, and tropical storms have contributed to widespread beach erosion. Some coastal regions resort to jetties, shore protection structures, and beach nourishment, whereby offshore or other nearby sediment sources are mined and added to the beach. These projects are costly, and their effectiveness must be closely monitored. This article investigates the ability of photogrammetry from an unoccupied aerial vehicle (UAV) to quantify geomorphic changes of the subaerial section to a newly nourished beach. On October 10, 2018, Hurricane Michael moved up the Gulf of Mexico past Tampa Bay, coinciding with an ongoing nourishment project at Indian Rocks Beach. We conducted UAV and ground surveys before and after the hurricane passage and compared point clouds and across-shore profiles at three locations. We compare observed erosional regimes to probabilistically forecast erosional regimes and found they were only minimally present, perhaps reflecting the recent sand additions. An average volume loss of similar to 31 m(3)/m was measured across the three studied sections.

As sea levels rise and temperatures increase, vegetation communities in tropical and sub-tropical coastal areas will be stressed; some will migrate northward and inland. The transition from coastal marshes and scrub–shrubs to woody mangroves is a fundamental change to coastal community structure and species composition. However, this transition will likely be episodic, complicating monitoring efforts, as mangrove advances are countered by dieback from increasingly impactful storms. Coastal habitat monitoring has traditionally been conducted through satellite and ground-based surveys. Here we investigate the use of UAV-LiDAR (unoccupied aerial vehicle–light detection and ranging) and multispectral photogrammetry to study a Florida coastal wetland. These data have higher resolution than satellite-derived data and are cheaper and faster to collect compared to crewed aircraft or ground surveys. We detected significant canopy change in the period between our survey (2020–2022) and a previous survey (2015), including loss at the scale of individual buttonwood trees (Conocarpus erectus), a woody mangrove associate. The UAV-derived data were collected to investigate the utility of simplified processing and data inputs for habitat classification and were validated with standard metrics and additional ground truth. UAV surveys combined with machine learning can streamline coastal habitat monitoring, facilitating repeat surveys to assess the effects of climate change and other change agents.

<p>The use of unoccupied aerial systems (UAS) in geoscience has dramatically improved our ability to collect data at high resolution, minimal cost, and in rapid response to sudden events. The wide range of sensor and platform configurations gives scientists great flexibility in survey design and data collection. Satellite remote sensing data has exceptional spatial coverage and continues to increase its data acquisition to meter-level resolution. UAS data can image to the cm-level resolution but lacks the same spatial coverage as satellite. By combining and comparing UAS data with satellite and ground-based remote sensing data we can utilize the different strengths of these systems. Here we demonstrate various UAS applications in high-resolution topographic change, land use classification, and sub-surface geological mapping. We use UAS payloads such as RTK georeferenced RGB and multispectral images, lidar, and magnetic sensors to image surface changes and sub-surface structures. We demonstrate the need for post-processing (PPK) high precision GNSS rover locations over utilizing only RTK position information.</p><p>Florida, USA, is home to rapidly changing beaches and wetlands, which are highly susceptible to our changing climate and destructive storm events. We show examples from beaches and wetlands in Pinellas County, Florida, USA where we have a) imaged the emergence and development of a barrier island, b) developed automated land use classification using photogrammetry and multispectral data, c) evaluated the impacts of a major hurricane event on a recently renourished beach. Pacaya Volcano, Guatemala, is an active volcano with frequent lava flows and historical flank collapse events. Using a combination of satellite DEMs, ground-based Terrestrial Radar Interferometry data, and UAS RGB SfM-photogrammetry, we have imaged recent lava flows in high-resolution showing details of lava flow levees and other structures. By comparing our data to pre-eruption satellite DEMs we can evaluate the volume and morphology of recently emplaced lava flows. In addition, we have collected magnetic data over recent lava flows that allows us to image the sub-surface structure of the lava flows and model lava flow properties. UASs are a powerful tool for remote sensing, geodetic, and geophysical data collection. They augment satellite and ground-based methodologies and by combining multidisciplinary data from these platforms we can image the earth in greater spatial and temporal detail than ever before.  </p>

Technological developments in the last few decades allow generation of increasingly high-resolution digital elevation models (DEMs), useful in many fields of Earth and environmental science, and especially for tectonic geomorphic studies. Combined with falling costs and the improved accuracy of geo-referencing using satellite geodetic tools based on Global Navigation Satellite Systems (GNSS), such as Global Positioning System (GPS), these developments have moved DEMs from the realm of computer equivalents of a topographic map to sophisticated tools for process understanding. Four techniques for the production of high-resolution DEMs are notable: light detection and ranging (LIDAR), interferometric synthetic aperture radar (InSAR), terrestrial radar interferometry (TRI), and structure from motion (SfM) photogrammetry. With the exception of TRI, restricted to ground-mounted platforms, the instrumentation can be hosted on satellites, piloted aircraft, or Unoccupied Aerial Vehicles (UAVs). Calibration with GNSS enables merging, or comparison of data sets acquired by different techniques, as well as change detection at the centimeter level.