Phenology monitoring observations have been traditionally linked to citizen science (CS) practices and limited to common species closely located to an observer’s residence. This is a modest approximation to cover the strong variability of the vegetation phenology across biomes and to improve the knowledge of vegetation changes as climate change impacts. Thankfully, remote sensing imagery (RS), especially data from high spatial and temporal resolution satellites such as Sentinel-2A and 2B, has the potential to become an essential dataset for in-situ approaches by guiding volunteers in achieving this aim. This study presents the methodological design carried out in Catalunya to harmonize phenological vegetation monitoring protocol between RS and CS making data acquired from both approaches interoperable. First, an Earth Observation Open Data Cube (ODC) containing Sentinel-2 radiometrically corrected data was created. The ODC offers a solution for storing big data products in an efficient way with low-cost hardware and provides easy access to data by indexing it. A selection of ~40 000 areas presenting a homogeneous cover common species of interest has been extracted from existing vegetation cartography. A time-series analysis of vegetation indices (greening and flowering) derived from the ODC has been performed for each area. A final subset of ~5 000 areas presenting a flowering amplitude signal > 10% and a greening-up amplitude signal > 20% was extracted and included in an interactive map browser available to CS observers so they are guided to interesting areas to report their phenophase and evaluate their correlation with RS data. This protocol, resulting from the PhenoTandem project (CSEOL - ESA) provides an innovative approach making in-situ observations interoperable with RS products and confirms a promising partnership for phenology monitoring.
One of the main concerns in adopting citizen science is data quality. Derived products inherit intrinsic limitations of the capture methodology as well as the uncertainties in observations. OpenStreetMap tools are designed to minimize uncertainties in positional accuracy by ensuring a good co-registration of the observations with imagery or direct use of GPS. When thematically annotating objects contributed by citizens, uncertainty increases. During the H2020 GroundTruth 2.0 project two land-cover products derived from OSM were analyzed; one created by the University of Heidelberg (http://osmlanduse.org) and another elaborated by University of Coimbra (https://vgi.uc.pt/vgi/osm/osm2lulc/). To be able to assess the quality of both maps, a third product derived from remote sensing was introduced as a reference map. In GroundTruth 2.0 a tool to show and compare maps as part of the MiraMon Map Browser was developed. The objective was to allow final users to auto-evaluate the quality of their region of interest. The confusion matrix has been used as a method to derive overall commission and omission estimators as well as the Kappa coefficient. Most of the discrepancies between OSM and remote sensing (RS) derived maps are related to different approaches used during data capturing. The data quality tool assesses the quality of individual observations exposed using the OGC standard and describes the quality in an interoperable approach based on QualityML.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.