Cattle and Ungulate Density, Eastern Norway

The Geovisualisation of Camera Trap Images

Introduction

It is estimated that 80% of the world’s data includes some kind of spatial aspect (MacEachren & Kraak 2001). Geovisualisation is a powerful tool capable of taking this spatial aspect and manipulating it for comprehensive data analysis and presentation (Dykes 2005). When used correctly, it allows the creator to present a topic of interest to a wide range of users from various backgrounds (Dykes 2005). Over the past two decades, the technologies involved in geovisualisation have expanded and become more interactive (Dykes 2005). Visualisations can now include temporal and 3D aspects with some geodata even being presented in virtual reality. This allows users to actively interact with visualisations, adding a level of personalisation that can’t be found in a static geodata report.

Geovisualisation is used often in ecology. A few examples include the spatial analysis of cattle disease (Dutra et al. 2022) and creating a 3D ecological network plan (Krisp 2006). Its applications are endless because there are thousands of ways to go about representing data depending on what you want to narrate to your audience. In this study, I will be exploring different ways of presenting an ecological project (details below) through interactive geovisualisation by using ArcGIS StoryMaps (Esri 2022).

LARGE and CarniForeGraze

First, it is necessary to provide a background on the project providing the data for geovisualisation so the user can understand where the data is coming from and the purpose of collecting such data.

LARGE is a research group focused on the sustainable management of large-bodied species (hence the name) such as moose and deer. Their research goals include;

  • Contributing to the knowledge surrounding these large-bodied animals and their interactions with other species, including humans.
  • Evaluating these species' management plans and the conflicts of interest involved.
  • Disseminating knowledge of these species to stakeholders, the scientific community, and the public.

CarniForeGraze is a sub-project of LARGE. This sub-project "studies the potential of using carnivore-exposed forests in southeastern Norway for livestock grazing in combination with forestry and large game hunting, while taking into account biodiversity of plants and pollinators". To be able to understand livestock and ungulate grazing and its effects on forest biodiversity the project acquired data on livestock and ungulate distribution using camera trap imagery.

We will explore the ways in which the images from the camera traps can be processed in order to come up with several visual representations that can be presented to a wide audience from different academic backgrounds.


Methods

Twenty-four camera traps were put up over the summer in Steinvik and Deset (see below) where they took pictures in 5-minute intervals for about three months. Pictures were uploaded and then looked through one by one to record the species present in each image. We captured a range of species, focusing on domesticated and wild ungulates (cattle, moose, red deer, roe deer), but including other wild species as well (bear, fox, birds, etc.) All locations were within 22km of one another.

Visualisations

Camera Trap Locations

The following visualisation allows the user to place the location of the camera traps in relation to the rest of Norway, allowing them to quantify the scale and the camera's respective locations in relation to one another. The images on the left-hand side add a visual stimulation as well as an insight into the raw data that was being collected from these camera traps. This gives the user an idea of how the individual animals were then counted from these images.

401

403

408

415

416

417

418

419

420

421

422

423

501

502

505

509

510

511

512

513

515

516

517

518

401

A sunny day at site 401.

403

Two researchers.

408

A staring moose in the undergrowth.

415

Image 1: A beautiful morning at site 415.

Image 2: A moose basking in the morning rays.

416

Image 1: A rainbow beyond the hills.

Image 2: Two researchers.

Image 3: A foggy morning.

417

Image 1: A red deer passing through.

Image 2: Two Researchers.

418

Image 1: A mother cow and its calf.

Image 2: A rather wet looking female moose.

Image 3: A bird of prey flying through the air.

419

Image 1: Cows resting and grazing.

Image 2: A cow with a collar grazing.

420

Two researchers.

421

Image 1: A magpie giving a wonderful display on top of the tree stump.

Image 2: Beautifully refracted light over sight 421.

422

Two researchers.

423

A mother moose and her calf passing through.

501

Image 1: A beautiful morning at site 501.

Image 2: Field workers creating the sapling plantation fences.

502

Image 1: Two researchers.

Image 2: A bird resting on a sapling.

Image 3: A moth flying towards the camera.

505

Image 1: A male red deer heading up the path.

Image 2: A herd of cattle resting and grazing.

Image 3: A foggy morning.

509

Image 1: Threecattle wandering through.

Image 2: A cow that appears to be scratching itself.

510

A female moose getting very close to the camera.

511

Image 1: A lone male moose.

Image 2: A staring moose calf.

512

Cattle resting and grazing.

513

A younger cow with ear tags passing through.

515

Image 1: Beautiful sun rays at site 515.

Image 2: A pretty sunset.

516

Image 1: A black grouse putting its head up.

Image 2: A moose peaking its head out with a possible calf.

517

Image 1: Cobwebs reflecting the sun's light.

Image 2: A field worker analysing sapling data.

518

Image 1: A red fox rests in the shrub.

Image 2: A moose calf passes through.

Image 3: An adult moose watching.

Image 4: Researcher.


Timeline Map

The timeline map below shows the sightings that occurred on each camera trap in one-day intervals. At the bottom of the map the timeline slider can be played and paused. The time interval on the slider can be altered by dragging the slider points. On the top left-hand side, the user can press the home button the re-centre the map and also take a screenshot using the image button. On the right-hand side, the legend of the map is displayed and can be toggled on and off. Layers can be filtered to show different species either by themselves or all together and the legend will change accordingly. The number of sightings for each camera is shown through the size of the circle. Details of numbers can be seen by clicking the legend section on the left tab. All layers are relative and show the same sizes for the same number.

Timeline Camera Trap Sightings - 02/06/2021 to 13/09/2021


Conclusive Map

The map below shows a conclusive insight on the sightings over the 3 months. Similarly to the timeline, the layers can be turned on and off to show each species by toggling the buttons in the top right-hand corner. The supplementary pie chart shows the percentages of species sighted (and numbers if scrolled over) and adjusts itself to the points that are within the map's extent.

Updated heatmap (copy)

Summary

The visualisations above create a useful representation of the camera trap data collected for the CarniForeGraze project. They are able to present the data in a way that can be understood by the general public and could serve a use in educating the public which is one of LARGE's project goals. As more data for the project is collected these visualisations could be used as a foundation for other features to be added on. For example, data on tree damage could be added to the map so it can be compared to animal density. More graphs like those seen on the conclusive map can be added and can interact with the data in different ways by the use of filters. Although the visualisations could be presented separately they work together well with the first being a stimulating picture-focused introduction, the second allowing you to see the data over time and the third being an overall conclusion of the data.

References

  1. Dutra, F., Navarro, M., Romero, A., Briano, C., Pereira, M. & Uzal F., A. (2022) Spatial and seasonal analysis and geovisualization of Fasciola hepatica–free bovine bacillary hemoglobinuria outbreaks in eastern Uruguay, 1999–2019. Preventive Veterinary Medicine. 199, Available: https://doi.org/10.1016/j.prevetmed.2021.105553
  2. Dykes, J., MacEachren, A.,M. & Kraak, M. (2005) Exploring Geovisualization. International Cartographic Association, 1-19. Available: https://doi.org/10.1016/B978-008044531-1/50419-X.
  3. Esri (2022) ArcGIS StoryMaps.
  4. Inland Norway University of Applied Sciences (2022) LARGE. Available: https://www.innlarge.no/
  5. Krisp, J., M. (2006) Geovisualization and knowledge discovery for decision-making in ecological network planning. Helsinki University of Technology: Publications in Cartography and Geoinformatics. Available: https://www.researchgate.net/publication/34664073_Geovisualization_and_knowledge_discovery_for_decision-making_in_ecological_network_planning
  6. MacEachren, A.,M. & Kraak, M. (2001) Research Challenges in Geovisualization. Cartography and Geographic Information Science, 28 (1), 3-12, Available: https://doi.org/10.1559/152304001782173970