Abstract:The Sentinel-2 data by European Space Agency were recently made available for free. Their technical features suggest synergies with Landsat-8 dataset by NASA (National Aeronautics and Space Administration), especially in the agriculture context were observations should be as dense as possible to give a rather complete description of macro-phenology of crops. In this work some preliminary results are presented concerning geometric and spectral consistency of the two compared datasets. Tests were performed specifically focusing on the agriculture-devoted part of Piemonte Region (NW Italy). Geometric consistencies of Sentinel-2 and Landsat-8 datasets were tested "absolutely" (in respect of a selected reference frame) and "relatively" (one in respect of the other) by selecting, respectively, 160 and 100 well distributed check points. Spectral differences affecting at-the-ground reflectance were tested after images calibration performed by dark object subtraction approach. A special focus was on differences affecting derivable NDVI and NDWI spectral indices, being the most widely used in the agriculture remote sensing application context. Results are encouraging and suggest that this approach can successfully enter the ordinary remote sensing-supported precision farming workflow.

Aerial LiDAR (Light Detection and Ranging) derived data are widely adopted for the study and characterization of forests. In particular, LiDAR derived-CHM (Canopy Height Model) has proved essential in identifying tree height variability and estimating many forest features such as biomass and wood volume. However, CHM quality may be affected by internal limits and anomalies caused by raw data (point cloud) processing (i.e., vertical errors), which are quite often disregarded by users, thus generating potentially erroneous results in their applications. In this work, an auto-consistent procedure for the fast evaluation of CHM accuracy has been developed based on the assessment of internal anomalies affecting CHM data obtained by differencing gridded DSM (Digital Surface Model) and DTM (Digital Terrain Model). To this purpose, a CHM was generated using the gridded DTMs and DSMs provided by the Cartographic Office of the Piemonte Region (north-western Italy). We estimated the local potential CHM error over the whole region, and demonstrated its strictly dependence on the terrain morphometry, particularly slope. The relationship between potential CHM error and slope was modeled separately for mountain, hill and flat terrain contexts, and used to produce a potential error map over the whole region. Our results showed that approximately 20% of the regional territory suffers from CHM uncertainty (in particular high elevation areas, including the treeline), though the majority of regional forest categories was affected by negligible CHM error. The potential consequences of CHM error in forest applications were evaluated, concluding that the tested LiDAR dataset provide a reliable basis for forest applications in most of the regional territory.

The Destination Earth (DestinE) European initiative has recently brought into the scientific community the concept of the Digital Twin (DT) applied to Earth Sciences. Within 2030, a very high precision digital model of the Earth, continuously fed and powered by Earth Observation (EO) data, will provide as many digital replicas (DTs) as the different domains of the earth sciences are. Considering that a DT is driven by use cases, depending on the selected application, the provided information has to change. It follows that, to achieve a reliable representation of the selected use case, a reasonable and complete a priori definition of the needed elements that DT must contain is mandatory. In this work, we define a possible theoretical framework for a future DT of the Italian Alpine glaciers, trying to define and describe all those information (both EO and in situ data) and relationships that necessarily have to enter the process as building blocks of the DT itself. Two main aspects of glaciers were considered and investigated: (i) the “metric quantification” of their spatial dynamics (achieved through measures) and (ii) the “qualitative (semantic) description” of their health status as definable through observations from domain experts. After the first identification of the building blocks, the work proceeds focusing on existing EO data sources providing their essential elements, with specific focus on open access high-resolution (HR) and very-high-resolution (VHR) images. This last issue considered two scales of analysis: local (single glacier) and regional (Italian Alps). Some considerations were furtherly reported about the expected glaciers-related applications enabled by the availability of a DT at regional level. Applications involving both metric and semantic information were considered and grouped in three main clusters: Glaciers Evolution Modelling (GEM), 4D Multi Reality, and Virtual Reality. Limitations were additionally explored, mainly related to the technical characteristics of available EO VHR open data and some conclusions provided.

Aerial discrete return LiDAR technology (ALS -Aerial Laser Scanner) is nowadays 6 widely used for forest characterization due to its high accuracy in measuring vertical 7 and horizontal forest structure. Random and systematic errors can occur and affect the 8 native point cloud, ultimately degrading ALS data accuracy, especially when adopting 9 datasets that were not natively designed for forest applications; a detailed understanding 10 of how uncertainty of ALS dataset could affect accuracy of derivable forest metrics (e.g. 11 tree height, stem diameter, basal area) is, in this case, required, looking for eventual 12 error biases that can be possibly modelled to improve final accuracy. In this work a low-13 density ALS dataset, originally acquired by the State of Minnesota (USA) for non-14 forestry related purposes (i.e., topographic mapping) was processed attempting to 15 characterize forest inventory parameters of the Cutfoot Sioux Experimental Forest 16 (north-central Minnesota, USA). Since accuracy of estimates strictly depends on the 17 applied species-specific dendrometric models a first required step was to map tree 18 species all over the forest. A rough classification, aiming at separating conifers from 19 broadleaves, was achieved by processing a Landsat 8 OLI scene. ALS-derived forest 20 metrics initially showed to greatly overestimate those measured at the ground in 230 21 plots. Oppositely, ALS-derived tree density was greatly underestimated, placing less 22 trees in the area. Aiming at reducing ALS measures uncertainty, trees belonging to the 23 dominated plane were removed from the ground dataset, assuming that they could not 24 properly be detected by low density ALS measures. Consequently, MAE

In the debate over global warming, treeline position is considered an important ecological indicator of climate change. Currently, analysis of upward treeline shift is often based on various spatial data processed by geomatic techniques. In this work, considering a selection of 31 reference papers, we assessed how the scientific community is using different methods to map treeline position and/or shifts using spatial datasets. We found that a significant number of published studies suffer from a low degree of awareness of processed data, which outcomes in potentially unreliable results that may compromise the validity of inference from the studies. Moreover, we propose an operational approach for easily incorporating consideration of spatial data quality, so as to improve reliability of results and better support ecological conclusions. Finally, we present a simulation of potential treeline vertical error for the Alpine region of Northern Italy, as driven by primary data quality.