Ash Dieback
A threat to roadside hedgerows: An assessment from air and space.
Dr Chloe Barnes
Head of Remote Sensing
Chloe joined the 2Excel Geo team in 2017, following the completion of her PhD in Remote Sensing. She is a domain expert in tree disease detection using spectral imaging and LiDAR techniques from airborne platforms and is an experienced data analyst.
The fungal pathogen Hymenoscyphus fraxineus is the causal agent for ash dieback which has been a significant problem in UK forestry since its introduction in 2012. As the disease has spread across the UK, the health and structural integrity of affected ash trees has become an increasing management concern. One environment particularly hard hit by ash dieback is roadside hedgerows where ash is often a dominant tree species.
In order to consider the potential for remote sensing imaging technologies to provide a means of detecting and monitoring decline in ash trees situated in hedgerows, 2Excel Geo investigated the potential applications of airborne and satellite platforms. The selection of the optimal remote sensing platform is largely dependent on spatial resolution and coverage requirements. Previous scientific studies have demonstrated use cases for both platforms in the monitoring of disease in forest environments (Abdel-Rahman et al., 2014; Zarco-Tejada et al., 2018).
The launch of the Sentinel-2a and 2b satellites in 2015 and 2017 has provided medium-resolution (10m+) multi-spectral imagery with high temporal resolution and a cost-effective opportunity to address challenges across all environmental sectors. To provide a comparison, airborne hyperspectral imagery was collected by 2Excel Geo at a resolution of 0.3 m for a hedgerow environment in Northamptonshire, England affected by ash dieback.
Figure 1
Sentinel-2 Image (Level 2A) for the area of Northamptonshire surrounding the hedgerow study site (21/06/2017).
Figure 1
Sentinel-2 Image (Level 2A) for the area of Northamptonshire surrounding the hedgerow study site (21/06/2017).
The aim of this particular case study was to compare the spatial resolution provided by the two platforms in relation to their potential applications in the assessment of ash trees in roadside hedgerows. Ground data regarding the location and crown dieback was acquired for 41 ash trees situated within the hedgerow.
Figure 2
True-colour aerial image of a section of the study site collected by 2Excel geo The green polygons represent ash trees identified in the hedgerow.
Figure 3
True-colour aerial image of a section of the study site collected by 2Excel geo The green polygons represent ash trees identified in the hedgerow..
In this type of hedgerow environment, tree crowns are often isolated from each other and situated amongst surfaces with very different spectral signatures such as bare earth and concrete. As a result, it is important that pixels selected for the assessment of ash dieback represent the reflectance values of the canopy of interest. The figure below demonstrates the difference in the spatial resolutions between the two remote sensing platforms. It is evident that the normalised difference vegetation index (NDVI) values derived from the Sentinel-2 image for the hedgerow ash trees were contaminated with the reflectance of the surrounding surface of the road and exposed soils.
Figure 3
Left; NDVI derived from Sentinel-2.
Right; NDVI derived from airborne imagery.
Figure 3
Left; NDVI derived from Sentinel-2. Right; NDVI derived from airborne imagery.
In more expansive forest areas, such as coniferous plantations, the impact of larger pixels is less problematic for disease monitoring as neighbouring trees of the same species will provide reflectance values more indicative of the vegetation characteristics. As a result, satellite platforms have previously been reported as an effective means of monitoring large-scale plantation environments in Canada and Scandinavia. Nevertheless, cloud cover presents a significant obstacle to the utilisation of optical data from satellites in certain regions of the world.
In the case of ash dieback assessment in roadside hedgerows in the UK, airborne or UAV systems which provide high-resolution remotely sensed datasets would be a more effective platform for data collection. The collection of imagery where single pixels can be solely associated with tree crowns will provide a more accurate means of characterising canopy condition.
