Three researchers from the University of Almería have won the prize for the best communication of the International Congress of Graphic Engineering, for a work in which they develop an efficient and precise methodology for carrying out forest inventories, with which it is possible to know how the forests over time.
This work have been make for Abderrahim Nemmaoui, Fernando José Aguilar and Manuel Ángel Aguilar and it is framed in the context of the development of two research projects whose main managers are professors José Ángel Aznar and Fernando José Aguilar, members of the research groups SEJ-579 ‘Environmental economics and natural resources’ and RNM-368 ‘Integrated Land Management and Spatial Information Technologies’respectively.
The first is a Challenge Project of the Junta de Andalucía called ‘Enabling interdisciplinary collaboration to foster Mediterranean forest sustainable management and socio-economic valuation (ECO2-FOREST)’. The second corresponds to the FEDER Aandalucía Operational Program, and its title is as follows: ‘Intervention strategies for an integrated and sustainable management of the Mediterranean forest based on an interdisciplinary analysis and its economic assessment (EGIS-FOREST)’.
What new possibilities does the winning research open?
The results obtained from the research that have been translated into the award-winning communication “make it possible to develop a efficient and sufficiently accurate methodology for fully digital forest inventories that allow us to know how Mediterranean forests evolve through the analysis of time series”. This is how the authors themselves have revealed it, abounding more with the idea that “These types of tools constitute the basis for monitoring and socio-economic assessment of the Mediterranean forest as an ecosystem that is especially vulnerable to the impact of Climate Change.proving to be essential for the development of a continuous strategy to support decision-making in what would be called ‘Integrated and Sustainable Management Policy for Forest Areas’”.
His starting point has been based on the fact that forests are among the most important ecosystems in the Mediterranean, due to the high biodiversity they house and the environmental services they offer. Two examples are biomass production and atmospheric CO2 fixation. “Knowledge of the ‘architecture’ of a forest is particularly key to estimating its environmental and economic value., as well as for the approach of interventions that contribute to its sustainable management”, they have stated. Traditionally, “the forest inventory is based on field sampling on small plots and statistical inference techniques allow large-scale data to be obtained from these initial data, although the procedure is slow, very expensive and, in general, too imprecise” .
Thus, “the irruption of data capture and analysis techniques based on modern Geomatics Engineering (drones, satellites, laser scanner) account for a large part of the research and development in the area of remote sensing of forests for carrying out inventories and supervision tasks on a small and medium scale”, allowing “precisely estimating various biophysical and structural parameters of the forest mass related to biomass associated, that is, tree height, crown diameter, stem height or normal diameter, among others”. Accurate modeling of the forest terrain surface, that is, “the reference topographic surface, or elevation ‘0’ of each tree, is necessary to normalize the 3D point clouds generated by digital stereo-photogrammetry from images ”.
They have explained that, “otherwise, the point cloud processing algorithm would not be able to determine, among other dendrometric variables, the height of the tree or its normal diameter (diameter of the trunk at 1.3 m height with respect to the base)”. In this sense, the availability of a Digital Terrain Model that is clearly appropriate “is crucial, as a starting point, for a correct semantic exploitation of the information contained in the 3D point cloud.” With this clear premise, this work has been developed with “a methodology based on stereo-photogrammetry from images taken from drones to generate these high-quality and precise Digital Terrain Models (DTMs) in Mediterranean forests. The images were taken in 31 square plots of Aleppo pine (Pinus halepensis Mill.) measuring 25 meters on each side located in the Sierra de María-Los Vélez Natural Park”.
The workflow has consisted of five successive phases: “Image orientation based on ground control points, generation of very dense 3D point clouds, automatic classification of terrain points, removal of outliers and interpolation of the DTM from the points classified as terrain by means of an algorithm designed for this purpose”. As a reference to determine the accuracy of the DTM obtained from images “a very precise DTM was used, generated by a terrestrial laser scanner (MDT-ELT)”. The results have shown that “the MDT-drones presented a reasonably low vertical systematic error of -9.38 centimeters, indicating that the MDT-dron tends to overestimate the z-elevation values of the ground provided by the MDT-ELT”. The vertical random error has been “only 4.68 centimeters, the average value for all plots in terms of standard deviation.” According to these results, “the MDT-dron can be considered very suitable as a reference surface in the development of digital inventories in Mediterranean forests.”
