More Like this

1. Transferability of the Deep Learning Mask R-CNN Model for Automated Mapping of Ice-Wedge Polygons in High-Resolution Satellite and UAV Images

   https://doi.org/10.3390/rs12071085  

   Zhang, Weixing ; Liljedahl, Anna K. ; Kanevskiy, Mikhail ; Epstein, Howard E. ; Jones, Benjamin M. ; Jorgenson, M. Torre ; Kent, Kelcy ( April 2020 , Remote Sensing)

   State-of-the-art deep learning technology has been successfully applied to relatively small selected areas of very high spatial resolution (0.15 and 0.25 m) optical aerial imagery acquired by a fixed-wing aircraft to automatically characterize ice-wedge polygons (IWPs) in the Arctic tundra. However, any mapping of IWPs at regional to continental scales requires images acquired on different sensor platforms (particularly satellite) and a refined understanding of the performance stability of the method across sensor platforms through reliable evaluation assessments. In this study, we examined the transferability of a deep learning Mask Region-Based Convolutional Neural Network (R-CNN) model for mapping IWPs in satellite remote sensing imagery (~0.5 m) covering 272 km2 and unmanned aerial vehicle (UAV) (0.02 m) imagery covering 0.32 km2. Multi-spectral images were obtained from the WorldView-2 satellite sensor and pan-sharpened to ~0.5 m, and a 20 mp CMOS sensor camera onboard a UAV, respectively. The training dataset included 25,489 and 6022 manually delineated IWPs from satellite and fixed-wing aircraft aerial imagery near the Arctic Coastal Plain, northern Alaska. Quantitative assessments showed that individual IWPs were correctly detected at up to 72% and 70%, and delineated at up to 73% and 68% F1 score accuracy levels for satellite and UAV images, respectively. Expert-based qualitative assessments showed that IWPs were correctly detected at good (40–60%) and excellent (80–100%) accuracy levels for satellite and UAV images, respectively, and delineated at excellent (80–100%) level for both images. We found that (1) regardless of spatial resolution and spectral bands, the deep learning Mask R-CNN model effectively mapped IWPs in both remote sensing satellite and UAV images; (2) the model achieved a better accuracy in detection with finer image resolution, such as UAV imagery, yet a better accuracy in delineation with coarser image resolution, such as satellite imagery; (3) increasing the number of training data with different resolutions between the training and actual application imagery does not necessarily result in better performance of the Mask R-CNN in IWPs mapping; (4) and overall, the model underestimates the total number of IWPs particularly in terms of disjoint/incomplete IWPs. 

   more » « less
2. Detecting Vegetation to Open Water Transitions in a Subtropical Wetland Landscape from Historical Panchromatic Aerial Photography and Multispectral Satellite Imagery

   https://doi.org/10.3390/rs14163976  

   Lamb, Lukas M. ; Gann, Daniel ; Velazquez, Jesse T. ; Troxler, Tiffany G. ( August 2022 , Remote Sensing)

   Over the last century, direct human modification has been a major driver of coastal wetland degradation, resulting in widespread losses of wetland vegetation and a transition to open water. High-resolution satellite imagery is widely available for monitoring changes in present-day wetlands; however, understanding the rates of wetland vegetation loss over the last century depends on the use of historical panchromatic aerial photographs. In this study, we compared manual image thresholding and an automated machine learning (ML) method in detecting wetland vegetation and open water from historical panchromatic photographs in the Florida Everglades, a subtropical wetland landscape. We compared the same classes delineated in the historical photographs to 2012 multispectral satellite imagery and assessed the accuracy of detecting vegetation loss over a 72 year timescale (1940 to 2012) for a range of minimum mapping units (MMUs). Overall, classification accuracies were >95% across the historical photographs and satellite imagery, regardless of the classification method and MMUs. We detected a 2.3–2.7 ha increase in open water pixels across all change maps (overall accuracies > 95%). Our analysis demonstrated that ML classification methods can be used to delineate wetland vegetation from open water in low-quality, panchromatic aerial photographs and that a combination of images with different resolutions is compatible with change detection. The study also highlights how evaluating a range of MMUs can identify the effect of scale on detection accuracy and change class estimates as well as in determining the most relevant scale of analysis for the process of interest. 

   more » « less
3. Mapping tundra ecosystem plant functional type cover, height and aboveground biomass in Alaska and northwest Canada using unmanned aerial vehicles, 2018-2019

   https://doi.org/10.18739/A2R785Q5B  

   Orndahl, Kathleen ( January 2022 , Arctic Data Center)

   Arctic landscapes are rapidly changing with climate warming. Vegetation communities are restructuring, which in turn impacts wildlife, permafrost, carbon cycling and climate feedbacks. Accurately monitoring vegetation change is thus crucial, but notable mismatches in scale occur between current field and satellite-based monitoring. Remote sensing from unmanned aerial vehicles (UAVs) has emerged as a bridge between field data and satellite imagery mapping. In this work we assess the viability of using high resolution UAV imagery (RGB and multispectral), along with UAV derived Structure from Motion (SfM) to predict cover, height and above-ground biomass of common Arctic plant functional types (PFTs) across a wide range of vegetation community types. We collected field data and UAV imagery from 45 sites across Alaska and northwest Canada. We then classified UAV imagery by PFT, estimated cover and height, and modeled biomass from UAV-derived volume estimates. Here we present datasets summarizing this data. 

   more » « less
4. Understanding the Effects of Optimal Combination of Spectral Bands on Deep Learning Model Predictions: A Case Study Based on Permafrost Tundra Landform Mapping Using High Resolution Multispectral Satellite Imagery

   https://doi.org/10.3390/jimaging6090097  

   Bhuiyan, Md Abul ; Witharana, Chandi ; Liljedahl, Anna K. ; Jones, Benjamin M. ; Daanen, Ronald ; Epstein, Howard E. ; Kent, Kelcy ; Griffin, Claire G. ; Agnew, Amber ( September 2020 , Journal of Imaging)

   Deep learning (DL) convolutional neural networks (CNNs) have been rapidly adapted in very high spatial resolution (VHSR) satellite image analysis. DLCNN-based computer visions (CV) applications primarily aim for everyday object detection from standard red, green, blue (RGB) imagery, while earth science remote sensing applications focus on geo object detection and classification from multispectral (MS) imagery. MS imagery includes RGB and narrow spectral channels from near- and/or middle-infrared regions of reflectance spectra. The central objective of this exploratory study is to understand to what degree MS band statistics govern DLCNN model predictions. We scaffold our analysis on a case study that uses Arctic tundra permafrost landform features called ice-wedge polygons (IWPs) as candidate geo objects. We choose Mask RCNN as the DLCNN architecture to detect IWPs from eight-band Worldview-02 VHSR satellite imagery. A systematic experiment was designed to understand the impact on choosing the optimal three-band combination in model prediction. We tasked five cohorts of three-band combinations coupled with statistical measures to gauge the spectral variability of input MS bands. The candidate scenes produced high model detection accuracies for the F1 score, ranging between 0.89 to 0.95, for two different band combinations (coastal blue, blue, green (1,2,3) and green, yellow, red (3,4,5)). The mapping workflow discerned the IWPs by exhibiting low random and systematic error in the order of 0.17–0.19 and 0.20–0.21, respectively, for band combinations (1,2,3). Results suggest that the prediction accuracy of the Mask-RCNN model is significantly influenced by the input MS bands. Overall, our findings accentuate the importance of considering the image statistics of input MS bands and careful selection of optimal bands for DLCNN predictions when DLCNN architectures are restricted to three spectral channels. 

   more » « less
5. Spectrally simplified approach for leveraging legacy geostationary oceanic observations

   https://doi.org/10.1364/AO.465491  

   Houskeeper, Henry F. ; Hooker, Stanford B. ; Cavanaugh, Kyle C. ( September 2022 , Applied Optics)

   The use of multispectral geostationary satellites to study aquatic ecosystems improves the temporal frequency of observations and mitigates cloud obstruction, but no operational capability presently exists for the coastal and inland waters of the United States. The Advanced Baseline Imager (ABI) on the current iteration of the Geostationary Operational Environmental Satellites, termed the$R$Series (GOES-R), however, provides sub-hourly imagery and the opportunity to overcome this deficit and to leverage a large repository of existing GOES-R aquatic observations. The fulfillment of this opportunity is assessed herein using a spectrally simplified, two-channel aquatic algorithm consistent with ABI wave bands to estimate the diffuse attenuation coefficient for photosynthetically available radiation,$K_d(\mathrm{P}\mathrm{A}\mathrm{R})$. First, anin situABI dataset was synthesized using a globally representative dataset of above- and in-water radiometric data products. Values of$K_d(\mathrm{P}\mathrm{A}\mathrm{R})$were estimated by fitting the ratio of the shortest and longest visible wave bands from thein situABI dataset to coincident,in situ$K_d(\mathrm{P}\mathrm{A}\mathrm{R})$data products. The algorithm was evaluated based on an iterative cross-validation analysis in which 80% of the dataset was randomly partitioned for fitting and the remaining 20% was used for validation. The iteration producing the median coefficient of determination ($R^2$) value (0.88) resulted in a root mean square difference of$0.319{\mathrm{m}}^{-<\#comment/>1}$, or 8.5% of the range in the validation dataset. Second, coincident mid-day images of central and southern California from ABI and from the Moderate Resolution Imaging Spectroradiometer (MODIS) were compared using Google Earth Engine (GEE). GEE default ABI reflectance values were adjusted based on a near infrared signal. Matchups between the ABI and MODIS imagery indicated similar spatial variability ($R^2=0.60$) between ABI adjusted blue-to-red reflectance ratio values and MODIS default diffuse attenuation coefficient for spectral downward irradiance at 490 nm,$K_d(490)$, values. This work demonstrates that if an operational capability to provide ABI aquatic data products was realized, the spectral configuration of ABI would potentially support a sub-hourly, visible aquatic data product that is applicable to water-mass tracing and physical oceanography research.

    

   more » « less