Assessing the sustainability of high-latitude rangeland grazing areas using high-resolution remote sensing imagery and spatial models of erosion


The overall aims of this project are to a) develop new approaches for quantifying erosion and land degradation in high-latitude environments from remotely-sensed data and b) develop predictive spatial models of soil erosion. Both of these aims will inform the sustainable use of this fragile resource into the future.

Land degradation involves a loss of ecosystem services. It is therefore important to be able to effectively monitor current land status and predict how that status may change in the future. High-latitude regions will be subject to rapid climate change and increased anthropogenic exploitation and it is unclear how eroded landscapes will be impacted. In Iceland, which has some of the most degraded landscapes in Europe (Arnalds, 2015), climatic warming and reductions in grazing intensity in recent decades would be expected to reduce the rate and intensity of soil erosion. However, in practice it has proved difficult to stop or reverse soil erosion, even when grazing is completely excluded. Greater understanding of the underlying biological and physical processes which control soil erosion patterns is needed in order to predict the response of Icelandic eroded landscapes to future global change.

The progression of soil erosion is readily assessed with remotely sensed data. Analyses of remote-sensing images have been used to test hypotheses about the underlying structural processes that generate patterns in vegetation cover (e.g. Scanlon et al, 2007). Vegetation patterns have also been used as an indicators of landscape resilience, and as early-warning signal of ecological collapse (Kefi et al., 2007). So far, these concepts have only been applied to low-latitude dryland environments, and it is not clear how widely applicable these ideas are to other environments, where the drivers and mechanisms of change are different.

Until recently, remotely-sensed vegetation data from satellites have only been available at relatively coarse resolutions (e.g. 10 m pixel size), limiting its use for understanding small-scale vegetation patterning. In Iceland this data has been used for simple quantification of total area of erosion, rather than to generate more nuanced understandings of landscape status. However, newer satellites such as Pleiades mean that higher resolution (up to 0.5 m pixel size) vegetation cover datasets and DEMS are now available. Pilot work in Iceland using unmanned aerial vehicles (UAV’s) has shown these data can be used these to generate a detailed understanding of spatial patterning in vegetation cover and soil loss. These newly available datasets provide a timely opportunity to enrich our understanding of how land degradation progresses in high-latitude areas. A preliminary spatial soil erosion model has been created – further development of this model through this project would give us a better understanding of how eroded areas emerge from underlying structural processes, and how they will respond to environmental change at scales relevant to land managers.

The research questions this project would address are:

1. Can we use spatial patterns of erosion (derived from remotely-sensed data at a variety of scales) as indicators of rangeland status for high-latitudes, particularly in Iceland?
2. What are the ecological processes that govern erosion patch expansion?
3. What is the likely trajectory and rate of soil erosion progression in Iceland in the future?

In addressing these questions, this project will be of great value to ensuring the sustainable use of rangeland grazing areas in high-latitude areas into the future.

Click on an image to expand

Image Captions

An eroded rangeland grazing area in northeastern Iceland


This study has three components 1) collecting and analysing remote-sensing imagery of eroded landscapes in Iceland 2) increasing our understanding of the processes that govern erosion patch growth and 3) the development of a spatial based model of erosion. The project will involve in total 5-weeks of fieldwork in Iceland for components 1 and 2.

Component 1: remote-sensing imagery
To assess the present, and recent, spatial patterning of eroded Icelandic rangeland environments this project will analyse remotely-sensed vegetation cover data across a range of spatial scales. We will use UAVs to collect very-high resolution data on vegetation cover (< 5 cm pixel size) for areas up to 0.25 km2. These datasets will be supplemented with high-resolution (0.5 m pixel size) satellite imagery ,allowing classification of key landscape metrics (e.g. patch-size frequency distribution, spatial pattern of erosion patches), over a range of ecological conditions and spatial scales. In addition DEM’s (e.g. the 2 m resolution Arctic DEM) will be used to assess changes in soil volume over time.

Component 2: understanding erosion patch growth
We know that there are biophysical feedbacks that promote erosion in certain parts of the landscape, typically near existing eroded areas. A more complete understanding of these processes is crucial for predicting future soil erosion. We will do this by undertaking small-scale ecological surveys and long-term (months-years) monitoring of environmental conditions (e.g. soil moisture and temperature) in actively eroding areas.

Component 3: spatial model of erosion
We need realistic spatial models of erosion to predict future changes. While conceptual models of erosion in Iceland exist (e.g. Dugmore et al., 2009; Barrio et al., 2018), there is a need for a spatially explicit model to provide quantitative predictions which can be used to aid management. Pilot work has created a simple model of erosion with distance-dependant feedbacks. This project will extend and improve this model.

Project Timeline

Year 1

In depth literature review, site selection, analysis of satellite remote sensing imagery, UAV training, Iceland fieldwork (2-weeks: UAV data collection and deployment of in-situ monitoring equipment)

Year 2

Iceland fieldwork (2-weeks) comprising further UAV survey, collection of data from in-situ monitoring equipment, further deployment of monitoring equipment. UAV data analysis and erosion model development

Year 3

Iceland fieldwork (1-week) to collect data from in-situ monitoring equipment. Data analysis of monitoring datasets, model outputs analysis, writing up thesis.

Year 3.5

Publication production and project dissemination.

& Skills

The student will, in collaboration with the supervisors, prepare a Training Needs Analysis and identify where he/she needs to add skills. These will be acquired in GradSkills courses given by the Centre for Academic and Professional Development (CAPOD) at St Andrews, in the IAPETUS2 training events and in external courses. During the course of the project the student will develop a wide range of advanced analytical skills. This training will ensure the student graduates with excellent statistical and GIS skills, which are highly valued across a range of research areas and within industry. Specific training will be provided for R-analysis, UAV piloting, the use of remote sensing software, GIS, spatial modelling using NetLogo and ecological survey techniques.

References & further reading

Arnalds, O. (2015) Soils of Iceland. Springer, Dordrecht.

Barrio IC, Hik DS, Thorsson J, Svavarsdottir K, Marteinsdóttir B and Jónsdóttir IS (2018) The sheep in wolf’s clothing? Recognizing threats for land degradation in Iceland using state-and-transition models. Land Degradation & Development. 29(6): 1714-1725: doi:10.1002/ldr.2978.

Dugmore, A.J., Gisladóttir, G., Simpson, I.A. and Newton, A. (2009) Conceptual models of 1200 years of Icelandic soil erosion reconstructed using tephrochronology. Journal of the North Atlantic, 2: 1-18.

Kefi S, Rietkerk M, Alados CL, Pueyo Y, Papanastasis VP, ElAich A, et al. (2007) Spatial vegetation patterns and imminent desertification in Mediterranean arid ecosystems. Nature. 449(7159): 213-217: doi:10.1038/nature06111.

Scanlon TM, Caylor KK, Levin SA and Rodriguez-Iturbe I (2007) Positive feedbacks promote power-law clustering of Kalahari vegetation. Nature. 449(7159): 209-U4: doi:10.1038/nature06060.

Further Information

Dr Richard Streeter,, 01334 463853

Apply Now