Publications

Recent publications on a range of topics, all using earth observation, remote sensing techniques to interpret the current understanding of the world.


Monitoring ash dieback (Hymenoscyphus fraxineus) in British forests using hyperspectral remote sensing
Fungal ash dieback (Hymenoscyphus fraxineus) is posing an imminent threat to forest health in Europe. Using airborne hyperspectral imagery trained against 422 tree crowns of known species and ash dieback severity, we built PLS-DA and RF models that classified individual tree crowns (ITCs) into five species (>90% OA) and ash crowns into three disease severity classes (77% OA) respectively. Dark pixel filtering was found to improve the accuracy of species (+6%) but not disease classification. By incorporating automatic ITC segmentation and the classification models, we further demonstrated how species and fungal ash dieback can be mapped at a region scale for forest management and epidemiological research. Chan (2020) Remote Sensing in Ecology and Conservation


Resilience of Spanish forests to recent droughts and climate change
Time-series of canopy greenness derived from satellite imagery can be analysed alongside environmental factors, species composition and management regimes, to better understand forest resilience to drought. In Spain, forests are on average greening despite drying trends. This resilience manifests in the short-term with native species activating drought tolerance and avoidance mechanisms observable from space (i.e. losing and gaining little greenness like chestnuts to losing and gaining a lot of greenness like maritime pines). The non-native eucalypt dominated forests reveal a low short-term resilience (i.e. do not recover enough after droughts) and hence have a higher percentage of declining pixels. Factors such as water balance, elevation, and protection status greatly influence these drought response patterns. Khoury (2020) Global Change Biology


Lateral meltwater transfer across an Antarctic ice shelf
We develop a semi-automated algorithm capable of tracking surface water bodies on Antarctic ice shelves, using a combination of Landsat 8 and Sentinel-2 imagery. In this paper, we apply our method to the Nivlisen Ice Shelf in the 2016-2017 melt season, and track changes in the geometry, area and volume of 1598 water bodies. We identify the greatest volume of surface melt ( 5.5×107  m3) on the 26th January 2017. On this day, 63 % of the total volume is held in two linear water bodies, which extend up to 27 km across the ice shelf surface. Dell (2020) The Cryosphere


Dynamics of a human‐modified tropical peat swamp forest revealed by repeat lidar surveys
In this study, two lidar surveys are compared to map forest biomass dynamics of PSF in Kalimantan, Indonesia. We found that historically logged forests were recovering biomass near old canals and railways used by the concessions. Lidar detected substantial illegal logging activity of logging canals were located beneath the canopy. Unexpectedly, rapid growth was also observed in intact forest that had not been logged. Carbon sequestration in above‐ground biomass may have offset roughly half the carbon efflux from peat oxidation. This study demonstrates the power of repeat lidar survey to map fine‐scale forest dynamics in remote areas, revealing previously unrecognized impacts of anthropogenic global change. Wedeux (2020) Global Change Biology


Combined InSAR and Terrestrial Structural Monitoring of Bridges
This paper examines advances in interferometric synthetic aperture radar (InSAR) satellite measurement technologies to understand their relevance, utilization, and limitations for bridge monitoring. Waterloo Bridge is presented as a case study to explore how InSAR data sets can be combined with traditional measurement techniques including sensors installed on the bridge and automated total stations. A novel approach to InSAR bridge monitoring was adopted by the installation of physical reflectors at key points of structural interest on the bridge, in order to supplement the bridge’s own reflection characteristics and ensure that the InSAR measurements could be directly compared and combined with in situ measurements. Selvakumaran (2020) IEEE Transactions on Geoscience and Remote Sensing


Supraglacial lake drainage at a fast-flowing Greenlandic outlet glacier
We present combined UAV and in situ records of a rapidly draining supraglacial lake in a fast-flowing sector of the Greenland Ice Sheet. Despite supraglacial lake drainage influencing ice sheet dynamics at a variety of scales, existing in situ studies have been conducted exclusively at the slower, less dynamic land-terminating sector. We describe the scale and extent of dynamic response in a marine-terminating system, and identify 1) spatially distributed behavior not previously observed in in situ studies, and 2) interannual variation unique to fast-flowing glaciers. We propose that many lakes thought to drain slowly are, in fact, draining rapidly via hydrofracture. As such, rapid drainage events, and their net impact on ice sheet dynamics, are being notably underestimated. Chudley (2019) PNAS.


Imaging spectroscopy reveals the effects of topography and logging on the leaf chemistry of tropical forest canopy trees
In this study we show that logged tropical forests have reduced leaf nutrient concentrations compared with old-growth forests and this becomes more pronounced as forests recover in stature. Our findings suggest rock-derived nutrients, such as phosphorus, in short supply in tropical forests on old soils, are depleted by as much as 30% by logging. This changes the concentration of these nutrients in leaves and may lead to shifts in species composition, and possibly reduced ecosystem function. To achieve landscape-scale maps of canopy nutrients, hyperspectral imagery was used to predict ground-based measurements taken directly from trees Swinfield (2019) Global Change Biology


A brave new world for archaeological survey: Automated machine learning-based potsherd detection using high-resolution drone imagery
This paper presents the first proof of concept for the automated recording of material culture dispersion across large areas using high resolution drone imagery, photogrammetry and a combination of machine learning and geospatial analysis that can be run using the Google Earth Engine geospatial cloud computing platform. The results show the potential of this technique, under appropriate field circumstances, to produce accurate distribution maps of individual potsherds opening a new horizon for the application of archaeological survey. The paper also discusses current limitations and future developments of this method. Garcia-Molsosa (2019) Journal of Archaeological Science


Mapping individual trees in tropical forests
Laser scanning has revolutionised forest ecology by providing high-resolution maps of forest structure over large spatial scales, but extracting information on individual trees remains a challenge. We describe a graph-based approach for delineating trees in dense forests which paves for remote sensing of the impacts of anthropogenic change on dense tropical forests. Williams (2019) IEEE


Beta-diversity of tropical forests using imaging spectroscopy
Why are tropical forests so biodiverse? Doesn’t survival of the fittest tell us that all but the most competitive species should be wiped out? We used remote sensing to map turnover of tree species across a tropical landscape. Composition varied with soil type and topography, indicating that species occupy different niches. Close-together patches were more alike in their species than those further apart, consistent with local dispersal of seeds. The study indicates that a combination of niches and “neutral” dispersal process help support the great diversity of species found in tropical forests. Bongalov (2019) Ecology Letters


High-accuracy UAV photogrammetry of ice sheet dynamics with no ground control
Unmanned Aerial Vehicles (UAVs) are increasingly common tools in the geosciences, but their use requires good ground control in order to make accurate georeferenced models. This is difficult in applications such as glaciology, where access to study sites can be hazardous. We show that a new technique utilising on-board GPS post-processing can match and even improve on ground-control-based methods, and, as a result, can produce accurate glacier velocity fields even on an inland ice sheet. Chudley (2019) The Cryosphere


Response of glacier flow and structure to proglacial lake development and climate at Fjallsjökull, south-east Iceland
Remote sensing can be used to gather detailed information on changes to a glaciers flow regime, structural architecture, frontal position, and terminal environment . This paper applies methods including structural mapping and feature tracking to Fjallsjökull, an outlet glacier in south-east Iceland. Dell (2019) Journal of Glaciology


Harmonising topographic & remotely sensed datasets, a reference dataset for shoreline and beach change analysis
There is value in harmonising coastal field-based topographic and remotely sensed datasets at local scales. Firstly, for the UK coast of North Norfolk, using open access UK Environment Agency datasets, shorelines are extracted from vertical aerial photography and validated against LiDaR (Light Detection and Ranging) and coastal topography surveys. Secondly, a standard methodology is provided for quantifying sediment volume change from spatially continuous LiDaR elevation datasets. As coastal systems are monitored at greater spatial resolution and temporal frequency there is an unprecedented opportunity to determine how and why coastal systems have changed in the past, with a view to informing future forecasting. Pollard (2019) Scientific Data


Dynamics of salt marsh margins are related to their three-dimensional functional form
Salt marsh margins represent the transition from an area too low in the tidal frame for vegetation to develop to an area high enough to be perennially vegetated. Analysis of UK Environment Agency annual vertical aerial photography between the Humber Estuary and the Thames Estuary, UK east caost shows that these margins can be statistically separated into three classes – ‘ramped’, ‘cliffed’ and ‘ridge-runnel’. Contrasting morphodynamic behaviours are associated with each margin type, providing a robust quantitative basis for a rapid evaluation of likely system dynamism that may be useful to conservation practitioners or site managers. Evans (2019) Earth Surface Processes and Landforms


Mapping tropical forest height with drones
Rapid advances in UAV technologies makes its possible to map forest canopies in 3D from photographs, using structure-from-motion techniques ( SfM ). However, it isn’t possible to see the forest floor through these photographs, limiting their usefulness for mapping forest structure. We developed a simple but effective way of estimating forest height and carbon stock from SfM. This could be a valuable for forest managers and restoration practitioners, providing the means to make rapid, low-cost surveys over hundreds of hectares without the need for LiDAR.  Swinfield (2019) Remote Sensing


The microclimate mapping challenge
Microclimates are often neglected in ecology and evolution. A key impediment has been the lack of spatial data to map microclimatic variation over large spatial scales and over time. Remote sensing is now offering opportunities to lift this technical barrier, by producing detailed and spatially continuous data-layers that can be used as explanatory variables to model microclimatic conditions over large spatial and temporal scales. We reviewed how these emerging technologies are advancing microclimate modelling and mapping, and highlight some of the opportunities they provide for ecology, conservation and climate change research. Zellweger (2019) Trends in Ecology and Evolution


Forest gain doesn’t stop forest fragmentation in China
Thanks to the rapid development of remote sensing technology, forest fragmentation maps are increasingly available to biodiversity studies. Using forest in China as a case study, we found that forest gains under China’s green policies have not entirely stopped fragmentation of natural forests. Road constructions and urbanisation are becoming the most influential drivers of forest fragmentation in recent decades. Liu (2019) Biological Reviews