An architecture and workflow are proposed and described for extracting geomorphic features from digital terrain model (DTM)-derived land surface parameters (LSPs) using deep learning (DL)-based semantic segmentation and integration of LSP calculations into the model architecture to allow for tensor- and graphics processing unit (GPU)-based computation within the DL framework. To characterize terrain patterns at multiple spatial scales, the input DTM can be generalized using Gaussian pyramids (GPs). The LSPs are then provided to the trainable component of the model. The workflow is explored using two examples: valley fill faces resulting from mountaintop removal surface coal mine reclamation, an example of anthropogenic geomorphic features, and sinkholes within a karst landscape, an example of natural geomorphic features. We compare three different, 'U'-shaped architectures as the trainable component of the architecture: a traditional, convolutional neural network (CNN)-based UNet, UNet with a ConvNeXt-based encoder and attention gates along the skip connections and UNet with a Mamba-based encoder and a CNN-based decoder. As a baseline, DL models are compared with a pixel-based random forest (RF) model using the same LSP feature space. We document improved performance in comparison with the pixel-based approach and minimal differences among the three DL architectures based on F1 scores. The inclusion of GPs did not have a large impact on predictive performance for either the RF- or DL-based models. Calculating LSPs as part of the model architecture did not greatly increase the computational complexity of the model. Depending on the model configuration, incorporating the LSP calculations into the model architecture increased the number of floating-point operations (FLOPs) and multiply-accumulate operations (MACs) by 2.8% to 17.6% and increased the saved model size by 1.2 MB. In practice, only DTM data, mapping extent(s) and example features need to be provided during the training process. Disk space usage is decreased by a factor of 6 when not using GPs and 31 when incorporating these multiscale representations. The workflow offers an efficient means to extract geomorphic features from DTMs and can support downstream modelling and research tasks. It has been made available in the R language via the geodl package and Python via the terrainseg package.

Prescribed fire is an established means to reduce wildfire risk and accomplish silvicultural and cultural objectives via controlled consumption of fuels. Terrestrial laser scanning (TLS) has been demonstrated as a simple and efficient means to quantify fuel and vegetation conditions in terms of three-dimensional forest structure and occupied volume at the plot-scale; however, there is currently a lack of studies that explore a repeat time series of TLS measurements for documenting fire treatment efficacy over time. We used a time-series approach to repeatedly scan 40 monitoring plots six times over a 4-year period (one pre-fire, one post-fire, and yearly scans over a 4-year period) within a burn unit in the New Jersey Pine Barrens, USA. Our TLS analysis evaluates metrics to quantify first-order effects on forest structure and subsequent structural changes and regrowth. Metrics include the percentage of all returns that were not classified as ground, as well as a volume-based, occlusion-adjusted percentage of non-occluded voxels with returns at the plot- and three strata-levels: canopy, tall shrub, and low shrub. We also made comparisons to other commonly used burn severity assessment methods, including the visual field-based composite burn index (CBI), as well as the Sentinel-2 Multispectral Instrument (MSI)-derived normalized difference vegetation index (NDVI) and normalized burn ratio (NBR). The TLS methods were most effective at characterizing changes in the low and tall shrub strata, which are commonly targeted during prescribed fire treatment. Post-fire change metrics were generally correlated with initial metric values and initial change in metric values, especially in the low and tall shrub strata where all Pearson and Spearman correlation coefficients were above 0.400 and as high as 0.955. We also document correlations among the TLS-based metrics and the initial change and rate of post-fire change in NDVI and NBR and the field- and visual-based composite burn index (CBI). Finally, we document that plot-level metrics can be aggregated to characterize entire burn units. Multi-temporal TLS compliments NBR- and CBI-based assessment and improves the quantification of prescribed fire efficacy for reducing the abundance of fuels in the low and tall shrub strata.

Accurate and scalable extraction of geomorphic features from LiDAR-derived terrain data is important for environmental monitoring, land use planning, and geohazard assessment. Recent deep learning advances have introduced state-space models, such as Mamba, which can capture broad spatial dependencies with linear complexity and offer an alternative to conventional convolutional neural network (CNN)- or Transformer-based approaches. This study evaluates Mamba-based architectures for geomorphic segmentation across three landform types: agricultural terraces, mine benches, and valley fill faces. Using high spatial resolution terrain derivatives, or land surface parameters (LSPs), as an input feature space, we compare a pure Mamba–UNet, a hybrid architecture with a Mamba-based encoder and CNN decoder, and a baseline CNN-based UNet with a ResNet-34 encoder, each trained with frozen and unfrozen encoders. Results indicate that no single architecture was best across all tasks; however, the hybrid Mamba-CNN architecture provided the most consistent performance overall, improving spatial continuity and interior feature recovery for tasks requiring broad spatial context, although its advantage over the baseline UNet was modest and task-dependent, and boundary-level analysis indicated that CNN-based configurations delineated feature edges most precisely, while the randomly initialized UNet remained highly competitive and achieved the strongest results for some tasks. Further, Mamba-based encoders converge quickly and maintain strong performance when frozen, whereas CNN-based UNets benefit more from encoder fine-tuning.

Anthropogenic reshaping of landscapes can increase landslide activity. One example of heavily modified, landslide-prone landscapes is the Appalachian region, which hosts weak geologic strata, steep slopes, a humid climate, geologically recent river incision, and large-scale landscape modifications resulting from surface mining. Surface mining conducted before the Surface Mine Control and Reclamation Act (SMCRA) of 1977 did not require reclamation to approximate original contour, so remnant highwalls and mine benches are common throughout Appalachia. This study investigated the occurrence of slope instability and failure beneath pre-law mine benches in the northern coalfields of West Virginia. Mine bench features were manually digitized based on interpretation of high-spatial-resolution digital terrain data and aerial orthophotography, and ancillary geospatial data (i.e., mine permit boundary and disturbance extent data sets). A subset of mine bench features, stratified by coal seam, was randomly selected for slope stability analysis. The lower margins of the randomly selected mine benches were differentiated into segments with and without evidence of slope failure or instability. Results suggest that local topographic slope and aspect were not strong determinants of slope instability beneath mine benches. However, the coal seam or unit that was mined was an important determining factor. The Pittsburgh seam, and to a lesser extent the Redstone seam, had a larger proportion of their lower margins showing evidence of slope failure than did other seams. This study highlights that local-scale geology controls the geomorphic response to human landscape modifications and suggests that proactive reclamation of surface coal mines might reduce slope-failure geohazards.

The application of deep learning for semantic segmentation has revolutionized image analysis, particularly in the geospatial and medical fields. UNet, an encoder-decoder architecture, has been suggested to be particularly effective. However, limitations such as small sample sizes and class imbalance in anthropogenic geomorphic feature extraction tasks have necessitated the exploration of advanced modifications to improve model performance. This study investigates a variety of architectural modifications to base UNet including replacing the rectified linear unit (ReLU) activation function with leaky ReLU or swish; incorporating residual connections within the encoder blocks, decoder blocks, and bottleneck; inserting squeeze and excitation modules into the encoder or attention gate modules along the skip connections; replacing the default bottleneck layer with one that incorporates dilated convolution; and using a MobileNetV2 architecture as an encoder backbone. Unique geomorphic datasets derived from high spatial resolution lidar data were used to evaluate the performance of these modified UNet architectures on the tasks of mapping agricultural terraces, mine benches, and valley fill faces. The results were further analyzed across varying training sample sizes (50, 100, 250, 500, and the full training set). Our results suggest that the incorporation of advanced modules can enhance segmentation performance, particularly in scenarios involving limited training data or complex geomorphic landscapes. However, differences were minimal when larger training set sizes were used (e.g., above 500 image chips) and the base UNet architecture was generally adequate. This research contributes valuable insights into the optimization of UNet-based models for anthropogenic geomorphic feature extraction and provides a foundation for future work aimed at improving the accuracy and efficiency of deep learning approaches in geospatial applications. We argue that one of the positive attributes of UNet is that it can be treated as a general framework that can easily be modified.

Semantic segmentation algorithms, such as UNet, that rely on convolutional neural network (CNN)-based architectures, due to their ability to capture local textures and spatial context, have shown promise for anthropogenic geomorphic feature extraction when using land surface parameters (LSPs) derived from digital terrain models (DTMs) as input predictor variables. However, the operationalization of these supervised classification methods is limited by a lack of large volumes of quality training data. This study explores the use of transfer learning, where information learned from another, and often much larger, dataset is used to potentially reduce the need for a large, problem-specific training dataset. Two anthropogenic geomorphic feature extraction problems are explored: the extraction of agricultural terraces and the mapping of surface coal mine reclamation-related valley fill faces. Light detection and ranging (LiDAR)-derived DTMs were used to generate LSPs. We developed custom transfer parameters by attempting to predict geomorphon-based landforms using a large dataset of digital terrain data provided by the United States Geological Survey’s 3D Elevation Program (3DEP). We also explored the use of pre-trained ImageNet parameters and initializing models using parameters learned from the other mapping task investigated. The geomorphon-based transfer learning resulted in the poorest performance while the ImageNet-based parameters generally improved performance in comparison to a random parameter initialization, even when the encoder was frozen or not trained. Transfer learning between the different geomorphic datasets offered minimal benefits. We suggest that pre-trained models developed using large, image-based datasets may be of value for anthropogenic geomorphic feature extraction from LSPs even given the data and task disparities. More specifically, ImageNet-based parameters should be considered as an initialization state for the encoder component of semantic segmentation architectures applied to anthropogenic geomorphic feature extraction even when using non-RGB image-based predictor variables, such as LSPs. The value of transfer learning between the different geomorphic mapping tasks may have been limited due to smaller sample sizes, which highlights the need for continued research in using unsupervised and semi-supervised learning methods, especially given the large volume of digital terrain data available, despite the lack of associated labels.

As the demand for geospatial analytics continues to grow, geographic information systems (GIS) professionals are needed to build, operate, and maintain GIS technologies, data, and software to provide geospatial insights for modern industries and organizations. To best train the next generation of GIS professionals, an understanding of qualifications and requirements of GIS positions is needed. Thus, this work analyzes 508 GIS positions, grouped by position type (analysts, developers, educators, managers, specialists, technicians) to provide insights on key pre-requisite requirements, such as education, experience, certifications, soft communication skills, programming skills, and knowledge of GIS or IT. In general, possession of a bachelor's degree in GIS, geography, or computer science, prior professional experience, and knowledge of GIS and IT software were common pre-requisites for most GIS roles. Soft communication skills were also frequently desired for GIS roles. We also found that some position requirements tended to vary by position type, such as manager and developer roles requiring on average 5 years or higher prior experience, while analyst, specialist, and technician roles had much lower experience and education requirements. Higher education institutions and GIS training programs should note the desired requirements for GIS position types and continue to refine programs and develop pathways for success for aspiring GIS professionals.