Models of stand volume and biomass estimation based on LiDAR data for the main forest types in Calabria (southern Italy)

Abstract

Models of stand volume and biomass estimation based on LiDAR data for the main forest types in Calabria (southern Italy). The AlForLab project is part of the Cluster MEA (Materials Energy Environment) addressed to the Calabria Region. Estimating the main dendrometric variables of Calabrian forests using models based on publicly available remote sensed data is one of the main purposes of the project. This paper describes the procedures used to develop several thematic maps (raster and vector) of timber volume and phytomass to be used in planning and management activities at both regional and forest property scale, as well as for felling plans, logging projects etc. We used public LiDAR data at medium-low resolution (1.6 pts m -2 ), acquired on about 90% of Calabrian territory in the frame of a national remote sensing programme of the Italian Ministry of the Environment. Field data from the second National Forest Inventory (INFC 2005) on 311 sample points were used for model calibration, as well as new field data acquired specifically for AlForLab project on 143 angle count samples. A series of regression models to predict volume and its corresponding aboveground biomass (dry and fresh weight) were developed and digital maps at different spatial resolutions were produced, as well as their estimate uncertainties. These models and their mapping products are also an important part of the new-establishing forest Decision Support System CFOR. The adopted models, though based on the same mathematical equation, have specific coefficients for different species and groups of species, according to a forest type classification system compatible with the fourth level of Corine Land Cover. In this way it is possible to apply the models without accessing more detailed forest type maps. All estimation methods and procedures are consistent with national forest inventory models, and with the other new tools proposed by AlForLab project to estimate timber volume, such as the regional tariffs and the field sampling inventory procedures. R 2 adjusted values (for models at the highest typological detail) are between 60% and 85%, whereas uncertainties of timber volume estimate (ESS%) range from 25% (for main forest species) up to 50% (for less spread forest types). All processing steps to produce digital maps were performed on open-source environment (R and QGIS).Keywords: ALS, LiDAR, Timber Volume, Forest Biomass, Estimation Models, CHM, AlForLab Received: Feb 09, 2017; Accepted: Apr 19, 2017; Published online: May 15, 2017 Citation: Scrinzi G, Floris A, Clementel F, Bernardini V, Chianucci F, Greco S, Michelini T, Penasa A, Puletti N, Rizzo M, Turco R, Corona P, 2017. Modelli di stima del volume e delle fitomasse del soprassuolo arboreo delle principali formazioni forestali della Calabria mediante dati LiDAR. Forest@ 14: 175-187 [online 2017-05-15]

UAV-DEMs for Small-Scale Flood Hazard Mapping

“…The surveying flights (scanning angle 25 • , flight height 1500-1800 m above ground level, pulse frequency 100 kHz) were carried out over a period of time from 2008 to 2014 (mainly in 2011). The geometric resolution of the pixel was 0.0001 • , corresponding to approximately 1 m. The native resolution of the LiDAR point cloud was averaged 1.6 pp/m 2 , recording up to 4 return echoes from the single sensor output pulse [53]. Despite the good accuracy of the LiDAR DEM, some interpolation errors, like triangle patterns in the sloping areas of the north-eastern and south-western part of the domain, were evident (Figure 1D).…”

Sustainable forest management and forest biomass estimates in Calabria (Southern Italy): a case study

“…Il calcolo dei volumi medi ad ettaro è stato effettuato applicando i modelli di stima proposti da Scrinzi et al (2017). Si tratta di nove modelli distinti per altrettante tipologie forestali caratteristiche del territorio calabrese, che stimano il volume legnoso medio a ettaro sulla base del valore medio del CHM, calcolato impiegando una matrice di esplorazione di 25 × 25 m (Hmean).…”

Detection and correction of LiDAR raster data from the Italian national remote sensing programme and production of a suitable CHM to forest volume estimation in Calabria (southern Italy)

“…L'effetto della presenza degli outlier nel CHM è stato esaminato confrontando i volumi medi all'ettaro e totali ottenuti applicando i modelli predittivi proposti da Scrinzi et al (2017) prima e dopo il processo di rimozione. Si tratta di nove modelli distinti per altrettante tipologie forestali proprie dell'ambiente boschivo calabrese che stimano il volume legnoso medio a ettaro sulla base del valore medio del CHM calcolato impiegando una matrice di esplorazione di 25 × 25 m (Hmean).…”