We recently discussed, in this post, just why forests are important to us, and the role they have to play in combating climate change. But what is the best way to measure forests? Every policy mechanism (and economic incentive) currently being proposed to combat deforestation is “performance-based”, meaning that robust mapping of forest properties is a fundamental component of the solution.
The inaugural Natural Capital Forum, held in Edinburgh before Christmas, supported an agreement that identified economic value of natural capital as the best vehicle for enabling the protection of key natural environments. This recognition of our reliance upon natural capital, and the need to protect them is also reflected by policy makers. At the COP19 in Warsaw there were significant advances in implementing financial mechanisms for protecting forests, known as REDD+. One of the most promising things to come out of the conference was the agreement to define a system for the monitoring, reporting, and verification of REDD+ projects. However, this is a complex problem which depends on a range of circumstances and requirements, but the need for robust monitoring systems which can feed into results based financing is clear. Remote sensing technologies have long been recognised as a primary vehicle for providing objective, practical, and cost-effective solutions for REDD+ monitoring. Indeed recent work by the University of Maryland, in partnership with Google and the US government, recently published the first freely-available, high-resolution global map of forest extent, loss and gain. Through the use of the Landsat satellite image series, they were able to produce a time series analysis of how forest cover has changed between 2000 and 2012, to a 30m spatial resolution.
So if we’re doing such a good job with satellite mapping of forests, why would you use laser scanning?
In short, satellites do a great job of some things, but airborne laser scanning is the most reliable way of properly mapping the forest.
A key area for the financing of REDD+ is the verifiable quantification of biomass. Traditionally this has been carried out through the use of field-based inventories; however these can be both expensive financially and in terms of time. The use of satellite data for this purpose has been used for a number of years; however there are limitations. The use of passive optical remote sensing satellites has primarily utilised coarse resolution mapping of 500m to 1km, which does not meet the accuracy required for REDD+ monitoring. And while very fine resolution optical remote sensing, such as Landsat or other commercial satellites, provides a measure of forest carbon stocks with a higher certainty, these systems all rely on indirect measurements of forest structure and various assumptions to estimate biomass.
Airborne laser scanning on the other hand provides direct measurements of the forest structure, giving actual measurements of tree height, and forest density, from which biomass can be determined. As they are active remote sensing systems, there is no reliance on solar illumination, which can be a significant limitation for passive optical remote sensing; where weather and time of day can have an impact on the measurements recorded.
The high spatial resolution provided by LiDAR also gives it a significant advantage over Synthetic Aperture Radar (SAR) providing much greater detail and thus more accurate results. In addition to this the use of full-waveform LiDAR means that the 3D structure of a forest can be mapped, providing detailed knowledge of the under-canopy which cannot be extracted by any other method. The greater knowledge of the vertical structure of the forest, as well as the horizontal structure, enables more accurate biomass estimates to be calculated. These characteristics are also highly beneficial for monitoring forest degradation; in particular the full-waveform overcomes the problem of canopy gaps closing which typically occurs 1 to 2 years after forest degradation has occurred. This problem currently limits the use of passive remote sensing which is only able to detect the top layer of the forest.
At Carbomap, we have also recognised the potential for multispectral LiDAR, which would negate the need for using a synergy of standard LiDAR and passive optical remote sensing; whereby the 3D structure of the forest is combined by the spectral signature, or “colour”. Specifically for REDD+ monitoring, it enables the use of LiDAR for increasing the wealth of information on many aspects of the MRV requirements. Such as forest degradation, and the extent to which forest fire risk and burnt area mapping can be achieved, aiding in the prevention and mitigation of fires which are detrimental to the success of REDD+ projects.
The nature of Airborne laser scanning means that some of the requirements of REDD+ monitoring cannot yet be met, such as near-real time monitoring of forest area change, deforestation, and land-use change. However the implementation of a spaceborne LiDAR system would overcome these challenges, extending the capabilities and applications of LiDAR for REDD+ monitoring. Unfortunately, there are currently no such systems operating, although the Carbomap team have developed a mission concept for a satellite laser scanner.
There has been an increasing acceptance from investors and government agencies that airborne LiDAR is the best solution for biomass mapping. The use of laser scanning for forest mapping and REDD+ monitoring, verification and reporting is growing as recognition of the benefits over other approaches is realised.
De Sy, V. Herold, M. Achard, F. Asner, G. Held, A. Kellndorfer, J. and Verbesselt, J. (2012) “Synergies of multiple remote sensing data sources for REDD+ monitoring”, Current Opinion in Environmental Sustainability, Vol. 4, 1-11, pp. 696-706