This Special Issue of the Journal of Maps is devoted to
highlighting contemporary examples of interdisciplinary collaborations between the arts and the geosciences (e.g. geomorphology, geology, Quaternary studies), with a specific focus upon the exploration of locations using, at least in part, some form of mapping. As previous
contributions to the journal have exemplified, mapping is essential for the exploration of locations, particularly by supplying visual representation to help with the characterisation of three core geographical concepts (Matthews & Herbert, 2008): space (e.g. distances, directions), place (e.g. boundaries, territories), and environment (e.g. biophysical characteristics).
FREE EPRINT: Testing and application of a model for snow redistribution (Snow_Blow) in the Ellsworth Mountains, Antarctica, Journal of Glaciology
Wind-driven snow redistribution can increase the spatial heterogeneity of snow accumulation on ice caps and ice sheets, and may prove crucial for the initiation and survival of glaciers in areas of marginal glaciation. We present a snowdrift model (Snow_Blow), which extends and improves the model of Purves et al. (1999). The model calculates spatial variations in relative snow accumulation that result from variations in topography, using a digital elevation model (DEM) and wind direction as inputs. Improvements include snow redistribution using a flux routing algorithm, DEM resolution independence and the addition of a slope curvature component. This paper tests Snow_Blow in Antarctica (a modern environment) and reveals its potential for application in palaeo-environmental settings, where input meteorological data are unavailable and difficult to estimate. Specifically, Snow_Blow is applied to the Ellsworth Mountains in West Antarctica where ablation is considered to be predominantly related to wind erosion processes. We find that Snow_Blow is able to replicate well the existing distribution of accumulating snow and snow erosion as recorded in and around Blue Ice Areas. Lastly, a variety of model parameters are tested, including depositional distance and erosion vs wind speed, to provide the most likely input parameters for palaeo-environmental reconstructions.
FREE EPRINT: Quantification of Hydrocarbon Abundance in Soils using Deep Learning with Dropout and Hyperspectral Data, Remote Sensing
Terrestrial hydrocarbon spills have the potential to cause significant soil degradation across large areas. Identification and remedial measures taken at an early stage are therefore important. Reflectance spectroscopy is a rapid remote sensing method that has proven capable of characterizing hydrocarbon-contaminated soils. In this paper, we develop a deep learning approach to estimate the amount of Hydrocarbon (HC) mixed with different soil samples using a three-term backpropagation algorithm with dropout. The dropout was used to avoid overfitting and reduce computational complexity. A Hyspex SWIR 384 m camera measured the reflectance of the samples obtained by mixing and homogenizing four different soil types with four different HC substances, respectively. The datasets were fed into the proposed deep learning neural network to quantify the amount of HCs in each dataset. Individual validation of all the dataset shows excellent prediction estimation of the HC content with an average mean square error of ~2.2×10-4. The results with remote sensed data captured by an airborne system validate the approach. This demonstrates that a deep learning approach coupled with hyperspectral imaging techniques can be used for rapid identification and estimation of HCs in soils, which could be useful in estimating the quantity of HC spills at an early stage.
FREE EPRINT: Assessment of low altitude UAS flight strategy on DEM accuracy, Earth Science Informatics
Soil erosion, rapid geomorphological change and vegetation degrada-
tion are major threats to the human and natural environment. Unmanned Aerial Systems (UAS) can be used as tools to provide detailed and accurate estimations of landscape change. The effect of flight strategy on the accuracy of UAS image data products, typically a digital surface model (DSM) and orthophoto, is unknown. Herein different flying altitudes (126-235 m) and area coverage orientations (N-S and SW-NE) are assessed in a semi-arid and medium-relief area where terraced and abandoned agricultural fields are heavily damaged by piping and gully erosion. The assessment was with respect to cell size, vertical and horizontal accuracy, absolute difference of DSM, and registration of recognizable landscape features. The results show increasing cell size (5-9 cm) with increasing altitude, and differences between elevation values (10-20 cm) for different flight directions. Vertical accuracy ranged 4-7 cm but showed no clear relationship with flight strategy, whilst horizontal error was stable (2-4 cm) for the different orthophotos. In all data sets, geomorphological features such as piping channels, rills and gullies and vegetation patches could be labeled by a technician. Finally, the datasets have been released in a public repository.
This paper outlines a collaborative project between a group of Fine Art and Geography students who helped develop and contribute to a conversation about recording ‘place’. Introducing methodologies from both disciplines, the project started from the premise of all environmental ‘recordings’ being ‘inputs’ and so questioned what could be defined as ‘data’ when encountering a location. Brunel’s Grand Entrance to the Thames Tunnel (London) provided the motivation for 10 objective and subjective ‘recordings’ which were subsequently distilled into a smaller subset and then used to produce a short film that was presented at an international conference. Important to the collaborative nature of the project were ongoing opportunities to share equipment, techniques, material and references across disciplines. It was an experiment to measure the potential for ‘mapping’ to capture physical and historical information, as well as embodied experience.
FREE EPRINT: Land inundation and cropping intensity influences on organic carbon in the agricultural soils of Bangladesh, Catena
Land inundation is a common occurrence in Bangladesh, mainly due to the presence of two major river systems -the Brahmaputra and the Ganges. Inundation influences land use and cropping intensity. However, there is little information on the influences of the extent of flooding and cropping intensity has on soil organic carbon (SOC),particularly at the landscape level. To investigate these influences, we collected 268 surface (0-30 cm) soil samples from 4 large sites within the two alluviums deposits (the Brahmaputra river and the Ganges river), on a regular grid (1600 m). The findings show that SOC levels are generally low, reflecting the intensity of agriculture and land management practices. SOC variability was higher across the medium high land (MHL) and medium low land (MLL) sites than in the high land (HL) and low land (LL) sites. The relatively low SOC levels and variability in the HL sites indicate soils here might have reached to equilibrium levels due to higher land use intensity. Topographically higher lands (HL and MHL), due to less of inundation, had higher cropping intensities and lower SOC’s than lower lands (MLL and LL), which had lower cropping intensities, as they remain inundated for longer periods of time. The findings clearly demonstrate the intrinsic influence of land inundation in driving cropping intensity, land management practices and SOC levels.
Creativity is one of those tropes that seems to do the rounds regularly in, well, creative circles. Almost by definition, it is levelled at the arts, in part because its base definition is along the lines of the ability to create. Withinthis context, cartography is well-poised because any map requires the cartographer to create a new, unrealised, graphic product.
OPEN ACCESS EPRINT: Demystifying academics to enhance university-business collaborations in environmental science
John K. Hillier, Geoffrey R. Saville, Mike J. Smith, Alister J. Scott, Emma K. Raven, Jonathon Gascoigne, Louise J. Slater, Nevil Quinn, Andreas Tsanakas, Claire Souch, Gregor C. Leckebusch, Neil Macdonald, Alice M. Milner, Jennifer Loxton13, Rebecca Wilebore, Alexandra Collins, Colin MacKechnie, Jaqui Tweddle, Sarah Moller, MacKenzie Dove, Harry Langford, and Jim Craig (2019)
challenge posed by a heavily time-constrained culture; specifically, tension exists between opportunities presented by working with business and non-optional duties (e.g. administration and teaching). Thus, to justify the time to work with business, such work must inspire curiosity and facilitate future novel science in order to mitigate its conflict with the overriding imperative for academics to publish. It must also provide evidence of real-world changes (i.e. impact), and ideally other reportable outcomes (e.g. official status as a business’ advisor), to feed back into the scientist’s performance appraisals. Indicatively, amid 20-50 key duties, typical full-time scientists may be able to free up to 0.5 day per week for work with business. Thus specific, pragmatic actions, including short-term and time-efficient steps, are proposed in a “user guide” to help initiate and nurture a long-term collaboration between an early- to mid-career environmental scientist and a practitioner in the insurance sector. These actions are mapped back to a tailored typology of impact and a newly created representative set of appraisal criteria to explain how they may be effective, mutually beneficial and overcome barriers. Throughout, the focus is on environmental science, with illustrative detail provided through the example of natural hazard risk modelling in the insurance sector. However, a new conceptual model of academics’ behaviour is developed, fusing perspectives from literature on academics’ motivations and performance assessment, which we propose is internationally applicable and transferable between sectors. Sector-specific details (e.g. list of relevant impacts and user guide) may serve as templates for how people may act differently to work more effectively together.
As a journal we notionally have two overlapping sets of “customers” - readers and authors. Authors provide the content whilst readers consume it. In a subscription funding model, readers pay for journal production, whilst in an open access (OA) model, authors pay. Somewhat uniquely in publishing, advertising plays a very limited part. And akin to commercial publishing, we have an overall journal editor (or Editor-in-Chief) and section editors (or Associate Editors).
FREE EPRINT: Hybrid Spectral Unmixing: Using Artificial Neural Networks for Linear/ Non-Linear Switching
Spectral unmixing is a key process in identifying spectral signature of materials and quantifying their spatial distribution over an image. The linear model is expected to provide acceptable results when two assumptions are satisfied: (1) The mixing process should occur at macroscopic level and (2) Photons must interact with single material before reaching the sensor. However, these assumptions do not always hold and more complex nonlinear models are required. This study proposes a new hybrid method for switching between linear and nonlinear spectral unmixing of hyperspectral data based on artificial neural networks. The neural networks was trained with parameters within a window of the pixel under consideration. These parameters are computed to represent the diversity of the neighboring pixels and are based on the Spectral Angular Distance, Covariance and a non linearity parameter. The endmembers were extracted using Vertex Component Analysis while the abundances were estimated using the method identified by the neural networks (Vertex Component Analysis, Fully Constraint Least Square Method, Polynomial Post Nonlinear Mixing Model or Generalized Bilinear Model). Results show that the hybrid method performs better than each of the individual techniques with high overall accuracy, while the abundance estimation error is significantly lower than that obtained using the individual methods. Experiments on both synthetic dataset and real hyperspectral images demonstrated that the proposed hybrid switch method is efficient for solving spectral unmixing of hyperspectral images as compared to individual algorithms.