As agroforestry systems have a high potential for food security, biodiversity and socio-economic sustainability and climate change adaptation and mitigation, they are currently being promoted or traditionally maintained in many regions of the world. In fact, agroforestry promotes multifunctional and resilient agriculture, with positive results in terms of ecosystem services. Due to the complexity of the tree-based agricultural systems an appropriate knowledge of its components, both in terms of spatial extent and functional relationships is fundamental. The mapping and monitoring of such complex systems are essential for a proper and sustainable management of resources as well as for planning of mitigation and adaptation measures against the rising environmental risks. The current estimation of agroforestry in European Union, according to LUCAS database (Eurostat 2015) is about 15.4 million ha, corresponding to about 3.6% of the territorial area and 8.8% of the utilised agricultural area (den Herder et al., 2017). Going down to the national level, 4,7% of total Italian area has been estimated as agroforestry, with main distribution in Central and Southern Italy. In this study, we aim to estimate agroforestry with more accuracy by exploiting the Google Earth Engine (GEE) platform. In order to achieve this target, we took a case study within a rural area in Central Italy, belonging to the "Bolsena Lake Bio-District". A bio-district is a civil local agreement, recognised by national and regional regulations, targeted to foster sustainability in rural areas by means of organic and high natural value farming systems. The Google Earth Engine cloud-computing platform (GEE) allows users a quick and seamless access to the standard satellite imagery without downloading the actual scenes, thus providing the means to build time series of indices counting hundreds of records in almost no time. GEE allows users to perform geospatial analysis from local to planetary scale based on Google's cloud infrastructure in a very short time, by accessing data from a large repository of publicly available geospatial dataset, including more than forty years of historical imagery, such as the entire Landsat archive as well as the complete Copernicus Sentinel archive and a variety of earth science-related datasets. In order to estimate and map agroforestry in the study area, we developed and tested an openly available GEE script, based on our previously studies for mapping Trees Outside Forest (TOF) in agroforestry landscapes (Chiocchini et al. 2019; Sarti et al. 2021). The workflow, in a nutshell, consists in singling out trees from a temporal series of images via optical indices thresholding, then extracting trees out of forest polygons (TOF) and classifying them according to their size and shape.
Exploiting the Google Earth Engine platform for mapping agroforestry in Italian rural landscape
Francesca Chiocchini;Marco Ciolfi;Maurizio Sarti;Marco Lauteri;Pierluigi Paris
2022
Abstract
As agroforestry systems have a high potential for food security, biodiversity and socio-economic sustainability and climate change adaptation and mitigation, they are currently being promoted or traditionally maintained in many regions of the world. In fact, agroforestry promotes multifunctional and resilient agriculture, with positive results in terms of ecosystem services. Due to the complexity of the tree-based agricultural systems an appropriate knowledge of its components, both in terms of spatial extent and functional relationships is fundamental. The mapping and monitoring of such complex systems are essential for a proper and sustainable management of resources as well as for planning of mitigation and adaptation measures against the rising environmental risks. The current estimation of agroforestry in European Union, according to LUCAS database (Eurostat 2015) is about 15.4 million ha, corresponding to about 3.6% of the territorial area and 8.8% of the utilised agricultural area (den Herder et al., 2017). Going down to the national level, 4,7% of total Italian area has been estimated as agroforestry, with main distribution in Central and Southern Italy. In this study, we aim to estimate agroforestry with more accuracy by exploiting the Google Earth Engine (GEE) platform. In order to achieve this target, we took a case study within a rural area in Central Italy, belonging to the "Bolsena Lake Bio-District". A bio-district is a civil local agreement, recognised by national and regional regulations, targeted to foster sustainability in rural areas by means of organic and high natural value farming systems. The Google Earth Engine cloud-computing platform (GEE) allows users a quick and seamless access to the standard satellite imagery without downloading the actual scenes, thus providing the means to build time series of indices counting hundreds of records in almost no time. GEE allows users to perform geospatial analysis from local to planetary scale based on Google's cloud infrastructure in a very short time, by accessing data from a large repository of publicly available geospatial dataset, including more than forty years of historical imagery, such as the entire Landsat archive as well as the complete Copernicus Sentinel archive and a variety of earth science-related datasets. In order to estimate and map agroforestry in the study area, we developed and tested an openly available GEE script, based on our previously studies for mapping Trees Outside Forest (TOF) in agroforestry landscapes (Chiocchini et al. 2019; Sarti et al. 2021). The workflow, in a nutshell, consists in singling out trees from a temporal series of images via optical indices thresholding, then extracting trees out of forest polygons (TOF) and classifying them according to their size and shape.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
