Urban areas & climate change Paolo Gamba University of Pavia
University of Pavia The UNIPV TLC & RS Lab is a dynamic research group developing in the past years many techniques devoted to the analysis and interpretation of remotely sensed data, with focus on urban areas and risk-related activities. The leader is Prof. Paolo Gamba, and selected reference past projects include: ESA IIMTS (2009-2010) Image Information Mining in long time data series FP6 HYPERINET (2007-2010) Training and research of young scientists in the field of hyperspectral imaging applied to vegetation, environmental models and land use mapping GEM GED4GEM (2010-2013) Exposure information extraction and storage by fusing existing database data and EO-derived information FP7 TOLOMEO (2011-2013) Definition of image processing tools for risk assessment using EO data processing for earthquake, flood and deforestation hazards FP7 MARSITE (2013-2015) Monitoring the Marmara Sea test site by means of multiple sensors (in situ and remote sensing) FP7 GSEXTANT (2013-2016) Security applications of remotely sensed data for scenarios outside Europe
Previous experience @ UNIPV UNIPV is member (zero funds) of the CCI-LC phase 2 consortium (lead by P. Defourny, Université Catholique de Louvain) UNPIV has on-going collaboration on urban land cover mapping in urban areas with Uni. Jena (CCI project) KTH Sweden (ESA DUE project) Univ. Nanjing & Chinese Academy of Sciences (P.R. China) for Dragon 4 The international consortium of GEO SB-04 task on Urban area monitoring (including Univ. Indiana State, USGS, National Observatory of Athens, JRC)
VHR Land Cover mapping Focus is on new R&D on ECVs started in CCI: Land Cover High resolution The idea is to obtain specialized data processing techniques, highly parallelizable and implemented into the ESA cloud, to obtain specific land cover types exploiting Sentinel 1 and 2 data sets Urban land cover Artificial structure Vegetation status in urban areas Beijing using Sentinel 1 and 2 and Multi-Scale Hierarchical Binary decision Trees Test Set 90% - Training Set 10% Overall Accuracy Average Accuracy Kappa HDT 85.66 78.30 0.83 Morph SVM SAR 73.58 64.97 0.69 Morph KNN SAR 69.89 60.15 0.65 Morph RF SAR 80.71 74.62 0.77 Morph SVM SAR 73.58 64.97 0.69 GLCM SVM SAR 63.47 49.97 0.57 GLCM KNN SAR 56.01 44.47 0.48 GLCM RF SAR 64.23 51.74 0.58 SVM OPT 66.80 55.78 0.61 KNN OPT 64.10 53.35 0.58 RF OPT 65.24 54.11 0.59 Morph SVM OPT 77.58 66.69 0.74 Morph KNN OPT 75.32 64.77 0.71 Morph RF OPT 80.52 69.12 0.77 GLCM SVM OPT 72.09 61.27 0.67 GLCM KNN OPT 60.28 51.34 0.53 GLCM RF OPT 73.63 61.70 0.69
Urban land cover using S1 & S2 Specialized procedure for urban extent mapping using Sentinel 1 Winner of the 2015 ESA Round Robin Tested in ESA EO4Urban project Validated for the whole China vs. Landsat extractions
Objectives and outcomes OBJECTIVES The overall objective of this research is to evaluate multitemporal Sentinel-1A SAR and Sentinel-2A MSI data for global urban services using innovative methods and algorithms, namely KTH-SEG, a novel object-based classification method for detailed urban land cover mapping, and KTH-Pavia Urban Extractor, a robust algorithm for urban extent extraction. Ten cities around the world in different geographical and environmental conditions are selected as study areas. Stockholm urban land cover mapping with KTH-SEG Beijing urban growth during 2010-2015 with KTH-Pavia Urban Extractor Sentinel-1 Sentinel-2 OUTCOMES This research and development is expected to produce a pilot global urban services demonstrator using multitemporal Sentinel-1A SAR and Sentinel-2A MSI data. The project will contribute to better understanding of the urban products and services required by end users development of novel and robust methods and algorithms for improved urban services to planners to support smart and sustainable urban development better understanding of the capacity of Sentinel-1A SAR and Sentinel-2A optical data for detailed urban land cover mapping and urbanization monitoring the goals and activities of GEO SB-04 Global Urban Observation and Information Task and the UN post-2015 sustainable development goals.
New Copernicus services At the global level, the outcome may be urban area global mapping and monitoring services, based on urban area extent maps. In disaster-prone regions featuring rapid urbanization, the service would be to provide up-todate and reliable urban land cover maps.
New Copernicus services At the city level, a local service will include three products: new built-up area maps, urban green structure maps and change maps.