Conference Agenda

As 2023 is set to be the warmest year on record, the accelerating impact of climate change is evident through rising sea levels and the increased frequency and intensity of extreme weather events. Cities concentrate people, buildings, and infrastructure in close proximity, making them particularly vulnerable to the effects of climate change. Presently, 56% of the world’s population live in cities, and this number is expected rise to nearly 70% by 2050. This urbanization trend underscores the urgency for concerted actions to mitigate the adverse effects of climate change on cities. Planners need timely and accurate information on urban land cover and its changing patterns, as well as vulnerability assessment to plan and implement mitigation strategies.

With synoptic view, global coverage and frequent revisits, satellite Earth observations (EO) have been offering unparalleled insights into the spatial and temporal dynamics of urban landscapes. By leveraging EO image time series and AI, the overall objective of this project is to enhance urban resilience and sustainability by developing novel approaches for urban mapping, urbanization monitoring, urban building and infrastructure monitoring and vulnerability assessment. The specific objectives are:

• to develop novel deep learning-based methods for urban land cover mapping using multi-source EO data;
• to develop novel deep learning-based methods for urban change detection in both 2D and 3D using SAR and optical dense time series;
• to develop the key techniques in geospatial ensemble learning of Earth Observation data for terrestrial surface vulnerability assessment;
• To develop methodologies and practical framework for EO-based urban safety monitoring and early warning on buildings and infrastructures.

The study areas include cities in Jing-Jin-Ji, Yangtze River Delta and Peral River Delta, and Hohhot-Baotou-Ordos-Yulin urban agglomerations, where the evolution of urbanization process will be monitored and terrestrial surface vulnerability will be analysed and assessed. The proposed methodologies include the development of deep learning models for urban land cover mapping, a multi-task deep learning method that can exploit EO image dense time series for urban change detection, a semi-supervised graph convolution algorithm for precise building change detection in densely populated urban areas, 3D semantic change detection by integrating multi-view and multi-modal imagery acquired by active and passive satellite sensors to produce reliable multi-temporal 3D semantic information,

The innovative aspects of this research include development of novel methodologies through the integration of EO and AI as well as contributing to resilient cities through vulnerability assessment and early warning. It is anticipated that detailed urban land cover information and their changes will be mapped detected in a timely and accurate manner. The urban change in 3D will be estimated to better understand urban density and environmental impact. This research is expected to contribute to 1) advance EO science, technology and applications beyond the state of the art, 2). timely and reliable updating of urban databases to support resilient and sustainable planning, 3) the monitoring objectives of the UN SDG11: make cities and human settlements inclusive, safe, resilient and sustainable and SDG13 Climate Action.

Urban environmental security refers to the protection of urban areas from environmental risks and threats that can adversely affect the well-being of their inhabitants, infrastructure, functions, and ecosystems. It encompasses various dimensions, among others climate change adaptation, natural resources management, pollution control, disaster risk reduction and urban resilience. In this research project, two of the above dimensions will be examined by means of Earth Observation methods and data:

1/ Urban Resilience, namely building the capacity of cities to withstand and recover from environmental shocks and stresses, including improving infrastructure to enhance resilience. To this end, the project delves into the natural and artificial dual water cycle mechanisms, land subsidence response, and regulation. By integrating remote sensing data with ground observations, transfer functions are optimized to identify cross-feed processes between water cycle elements and land subsidence. Employing technical methods like trend tests and cross wavelet analysis, change trends and response relationships are analyzed. Ground response patterns driven by changes in groundwater levels will be recognized and collaborative attribution analysis using water balance models and machine learning will reveal the coupling evolution mechanism of dual water cycles and land subsidence, informing for the regulation thresholds and settlement schemes for sustainable urban development.

2/ Climate change adaptation, namely developing strategies to mitigate and adapt to the impacts of climate change, such as extreme weather events, urban heat and heatwaves, which pose significant risks to urban populations, environment, and infrastructure. To this end, the project will exploit the link between the state of the urban thermal environment and the 3Us, namely urban functions, urban form and urban fabric. Advanced downscaling methods and machine learning to recognize temporal patterns of urban heat and thereafter forecast thermal risk dynamics using a specialized Recurrent Neural Network (RNN). Additionally, the daytime surface Park Cool Island effect will be examined, analyzing factors such as size and shape of urban parks to assess the mitigation and adaptation potential of urban parks to climate change. Finally, the impact of urban building morphology on air ventilation patterns is assessed through atmospheric circulation analysis, Earth observation, GIS, and numerical modeling, contributing to optimization in urban planning.

The Pan-Third Pole region covers 20 million km², encompassing the Tibetan Plateau, Pamir, Hindu Kush, Iran Plateau, the Caucasians, the Carpathians, etc. and is home to over 3 billion people. Climate change is expected to dramatically impact the water and energy as well as carbon cycles and exchanges in this region and consequently alter the water resources, food security, energy transition and ecosystems as well as other related societal challenges. Monitoring and modelling climate change in this region reflect key societal issues and contribute to the science component to other international initiatives, e.g. UN sustainable development goals (SDG), GEO societal benefit areas and the ESA EO science for society strategy. The objectives of this DTE-CLIMATE project are: (1) to advance the process understanding of the interactions between the Asian monsoon, the plateau surface (including its permafrost and lakes) and the Tibetan plateau atmosphere in terms of water, energy and carbon budgets and to extend such understanding to the whole Eurasia by means of the Digital Twin Earth (DTE) approach; (2) to assess and monitor changes in cryosphere and hydrosphere; and (3) to model and predict climate change impacts on water resources and ecosystems in the Pan-Third Pole Environment. The huge amount of data and diverse models developed in previous projects will be integrated following the DTE concept to serve various application purposes, in particular, this will be applied in Europe for monitoring climate extremes (e.g. drought and flood) and assessment of ecosystem resilience to climate change. A core innovation of the DTE-CLIMATE project is to verify or falsify recent climate change hypotheses and projections of the changes in water and carbon cycles and their impacts on water resources and ecosystems with quantitative satellite observations via a DTE approach.

Method: We will use Earth observation, in-situ measurements and modelling to advance process understanding relevant to monsoon scale predictions, and to integrate such knowledge into a DTE for pan-TPE and Eurasia to explain different physical links and scenarios that cannot be observed directly. The diverse processes and data will be integrated in a DTE context by data assimilation to integrate satellite observations with model simulated states for large scale monitoring and assessment purposes. We will seek to utilize cloud and artificial intelligence (AI) tools and facilities that are under development by ESA and the EU Destination Earth.

Deliverables: The deliverables will be scientific outputs in terms of peer reviewed journal publications, PhD theses and data sets in terms of novel data records and modelling tools of essential climate variables for the quantification of water, energy and carbon cycle dynamics. These assets will be integrated in the framework of DTE for easy applications.

This study, initiated by the two groups in 2009, focuses on the development of corrosion/soiling models using ground-based data. Building upon these findings, the research aims to develop corrosion/soiling models utilizing satellite data under the DRAGON 3 and 4 projects.

The primary objective of this project is to utilize satellite environmental and air pollution data to estimate material deterioration and soiling levels and to develop innovative tools known as Satellite Sensed Data Dose-Response Functions (SSD-DRFs). The project specifically focuses on selected monuments in China and Europe and consists of three main tasks: 1) Development of deterioration and soiling indices, 2) Creation of new SSD-DRFs for copper and stainless steel, and 3) Seismicity monitoring in culturally significant areas using satellite observations.

The main satellite data sources include Sentinel-5p, Sentinel-4, FY-3A/B/C/D/E, and CSES. The methodology for each task is as follows:

- Task 1 (Months 0-36): Utilize SSD-DRFs to generate the Limestone Deterioration Index (LDI) and Soiling Index (SI) for selected monuments. Historical satellite data will be used to create databases of degradation and soiling time series.

- Task 2 (Months 0-48): Develop new SSD-DRFs for copper and stainless steel using Artificial Intelligence and Machine Learning Techniques. Compile databases of environmental, air pollution, and experimental deterioration data for these materials.

- Task 3 (Months 0-48): Monitor seismic activity in archaeological areas using satellite observations. Particular focus will be given to sites such as the Archaeological site of Delphi in Europe and the Temple and Cemetery of Confucius in China.

The primary deliverables of this project will consist of:

- Compilation of databases containing degradation and soiling time series values for selected monuments, copper, and stainless steel.

- Development of new SSD-DRFs for copper and stainless-steel materials using experimental and satellite data.

- Creation of a prototype seismicity database for archaeological sites to aid in the protection of cultural heritage sites.

- Establishment of the Limestone Deterioration Index and Soiling Index for the selected monuments to offer insights into the atmospheric impact on the materials.