Search for a project

Monitoring Forest Ecosystem Services with Multi-Sensor Earth Observations

Dr Guy Ziv (SoG), Dr David Galbraith (SoG), Darren Moseley (Forest Research), Juan Suarez (Forest Research)

Project partner(s): Forest Research (CASE)

Contact email: g.ziv@leeds.ac.uk

Project Summary

Forestry in Britain has undergone major shift in focus over the last century from timber production as the main objective to delivery of multiple ecosystem services such as carbon sequestration, soil protection, maintaining biodiversity, improving water quality and providing recreational opportunities. National regulations and policies, as well as the European Habitats Directive, the European Water Framework Directive, and the Convention on Biological Diversity Aichi 2020 targets, are all calling for a cost-effective monitoring system to track the spatial and temporal evolution of forests’ structure, functions and ecosystem services (ESS). The recent advances in Remote Sensing (RS), in particular the availability of free satellite data (e.g. Landsat and Sentinel-2), the dramatic drop in the cost of UAVs (drones) carrying thermal, LiDAR and hyperspectral sensors, and recent cloud-processing capacity (e.g. Google Earth Engine) open the door for RS-based monitoring of forest ESS. However, combining multi-sensor and multi-temporal RS data and mapping the diversity of forest services is still at its infancy. This project aims to decipher such relationships using in-situ data (ecological, management and ESS) gathered in and around one of the UK three Research Forests (Queen Elizabeth Forest Park - QEFP) and a large RS dataset (satellite and aerial imagery, LiDAR, thermal and hyperspectral UAV data), and to establish a prototype RS-based monitoring system for UK forests.

Background

Forests provide numerous ecosystem services, including provisioning (e.g. timber, berries, mushrooms, game species), regulating (e.g. carbon sequestration, flood risk reduction, improving water quality) and cultural (e.g. recreation). The provision of these ESS depends on the landscape composition (i.e. area occupied by different forest types as well as agricultural land and urban areas) but also its configuration1–3. However, it is well-known that land use and land cover alone are insufficient to identify the spatial pattern of ESS4, with several recent papers demonstrating how ecosystem properties, ecological functions and ESS correlate with tree species richness5,  functional traits diversity6,7 or a combination thereof8. Furthermore, management history has long-lasting effect on the flow of ESS9,10. As an example, a mosaic of irregular-aged stands of short rotation conifers would differ from mixed broadleaved trees managed using a shelterwood in species diversity and vertical structure, resulting in different levels of biomass harvest, early and late successional biodiversity, and recreational value. While there has been increasing interest in monitoring both structural and functional traits from space11,12, major gaps exists in linking multi-sensor data, in particular combining hyperspectral and LiDAR metrics, to ecosystem properties, ecological functions and ESS.

Research Objectives

This PhD project will consider the following three challenges:

  1. Linking (a) forest management information to (b) species and functional traits diversity and forest structure, (c) ecosystem properties and ecological functioning and (d) provision of various ecosystem services both within and on agricultural fields nearby QEFP
  2. Testing data and algorithms (e.g. machine learning, supervised and unsupervised classifiers) to link multi-sensor and multi-temporal RS data to (a) forest management; (b) species and functional traits metrics; (c) ecosystem services indicators
  3. Building a prototype online ecosystem services monitoring system for the QEFP

Data and Methods

The project will build upon data available from multiple sources such as Forestry Commission sub-compartment silviculutral management data, national inventory of woodland and trees, Scottish semi-natural woodland inventory and planting data, and the TRY database13. Furthermore, Forest Research collected in-situ data on carbon stocks, timber production, biodiversity, recreation, flood mitigation, water quality and air quality in the QEFP. Fieldwork may be necessary to complete missing information, and measure ESS provisioning on agricultural land near QEFP. There is also a number of RS data already available for this analysis, including 20+ years of satellite (Landsat, Seninel-2) and aerial imagery and 4 LiDAR surveys conducted between 2002-2012. Forest Research are collecting thermal and hyperspectral UAV data, and the student will take part in that campaign.

Using all this in-situ and RS data in geospatial regression analysis will help link together management, structure, function and ESS. It will also allow to adapt and improve methods to monitor the key factors with RS data, using e.g. Google Earth Engine. The final objective will be met as part of a 3 months placement in Forest Research Northern Research Station (NRS) stationed near Roslin, Scotland. This will require some interviews with knowledge end users to define their needs and help translate the outcomes and algorithms into an online Google App Engine ESS monitoring prototype.

Queen Elizabeth Forest Park

Queen Elizabeth Forest Park (QEFP) is a Research Forest located near Aberfoyle at the boundary of the Scottish highlands and lowlands, and lies within the Loch Lomond and the Trossachs National Park. The park covers 67,000 hectares and includes areas of native woodland, productive forest, water courses and areas of open space. It is managed by Forestry Commission Scotland to deliver multiple objectives, including recreation, biodiversity and timber production - whilst maintaining the character of landscape. The QEFP receives over 1 million visitors each year, incorporates a large complex habitat network, and supports the local and national timber industry. Forest managers at QEFP are committed to managing their forests to be resilient to climate change and have already begun making changes to the forest and how it is managed.

Student profile

This project need a highly motivated and enthusiastic student, who should have, or expect to receive, a first class BSc degree, or a distinction at Masters level, in an appropriate discipline. Preferably, they should have interest and experience in as many of the following topics: GIS (raster and vector analysis), ecosystem services assessment (mostly physical methods), statistical analysis (generalized linear regressions and/or mixed models using R or MATLAB), computer scripting (e.g. Python, Javascript etc.), conducting fieldwork (e.g. collecting transects or quadrats data), forest management (knowledge of silvicultural practices), and remote sensing analysis (e.g. manipulating multispectral data, using vegetation indices). The project will include significant fieldwork component, and necessitate frequent travel to/from QEFP and NRS.

Skills and training

Training will be provided in advanced geospatial analysis, including cloud-computing (Google Earth Engine platform). Support will be given to improve the student’s skills in literature review, conducting fieldwork, computer scripting, statistical analysis and data interpretation, remote sensing data manipulation, presenting results, and scientific writing. The student will join a very active and diverse research group of the Ecology and Global Change cluster, and receive constructive feedback from peers in cluster meetings and university postgraduate research days. An additional important part of the research training will be to attend national and international conferences to present results and gain feedback. The student will be encouraged to write and submit papers for publication during the project. Further training in project-specific field and office-based techniques will be provided.

CASE partnership

This project will be a CASE partnership with the Forest Research Northern Research Station (NRS) stationed near Roslin, Scotland.  Forest Research will provide a £1k salary top-up per year to the student, as well as provide supervision time and a T&S for travel to NRS. As part of the studentship, the student will spend 3 months placement in NRS. Forest Research will also support he project with in-kind supervision from Land Use and Ecosystem Services (LUES) Research Group (Moseley) and Remote Sensing Applications Programme (Suarez).

Enquiries

Informal enquiries should be directed to Dr. Guy Ziv (g.ziv@leeds.ac.uk).

Further Reading

1.        Mitchell, M. G. E. et al. Reframing landscape fragmentation’s effects on ecosystem services. Trends Ecol. Evol. 30, 190–198 (2015).

2.        Verhagen, W. et al. Effects of landscape configuration on mapping ecosystem service capacity: a review of evidence and a case study in Scotland. Landsc. Ecol. 31, 1457–1479 (2016).

3.        Lamy, T., Liss, K. N., Gonzalez, A. & Bennett, E. M. Landscape structure affects the provision of multiple ecosystem services. Environ. Res. Lett. 11, 124017 (2016).

4.        Eigenbrod, F. et al. The impact of proxy-based methods on mapping the distribution of ecosystem services. J. Appl. Ecol. 47, 377–385 (2010).

5.        Gamfeldt, L. et al. Higher levels of multiple ecosystem services are found in forests with more tree species. Nat. Commun. 4, 1340 (2013).

6.        Paquette, A. & Messier, C. The effect of biodiversity on tree productivity: From temperate to boreal forests. Glob. Ecol. Biogeogr. 20, 170–180 (2011).

7.        Lavorel, S. et al. Using plant functional traits to understand the landscape distribution of multiple ecosystem services. J. Ecol. 99, 135–147 (2011).

8.        Díaz, S. et al. Incorporating plant functional diversity effects in ecosystem service assessments. Proc. Natl. Acad. Sci. U. S. A. 104, 20684–20689 (2007).

9.        Becagli, C. et al. Monitoring managed forest structure at the compartment-level under different silvicultural heritages: An exploratory data analysis in Italy. J. Sustain. For. 35, 234–250 (2016).

10.      Yu, X. et al. Modelling long-term water yield effects of forest management in a Norway spruce forest. Hydrol. Sci. J. 60, 174–191 (2015).

11.      Jetz, W. et al. Monitoring plant functional diversity from space. Nat. Plants 2, 16024 (2016).

12.      Abelleira Martínez, O. J. et al. Scaling up functional traits for ecosystem services with remote sensing: concepts and methods. Ecol. Evol. 6, 4359–4371 (2016).

13.      Kattge, J. et al. TRY - a global database of plant traits. Glob. Chang. Biol. 17, 2905–2935 (2011).

Related undergraduate subjects:

  • Biodiversity
  • Ecology
  • Natural resource management
  • Remote sensing
  • Sustainability and environmental management