Status

The project is ongoing.

Summary

Satellite-based synthetic aperture radar (SAR) interferometry has become an established remote sensing technique to retrieve geophysical parameters such as surface topography and monitoring of ground displacements (deformation). SAR is particularly well suited for monitoring tasks since radar images can be acquired irrespective of day/night-time and cloud conditions. This project aims at advancing SAR interferometric methods for improved identification of critical landslide and rockfall, and other mass movements in mountainous areas in general, and preemptive assessment of triggers leading to slope failures. In this context, this project aims to build a transboundary research partnership between relevant experts from Switzerland and Pakistan. These experts, who are authors of the proposal itself, have set the agenda together on how to achieve the aforementioned goals and overcome the challenges associated with SAR interferometry in mountainous regions in general. Firstly, there is a need to develop improved and updated high resolution landslide susceptibility maps. While Switzerland has already well defined hazard maps, Pakistan is lagging behind. Nonetheless, both countries require continued monitoring at regional scales which is effectively possible only via spaceborne remote sensing methods like SAR interferometry. Moreover, an enhancement of the quality and spatial coverage of spaceborne SAR interferometry-based deformation measurements is needed, especially in rugged mountainous areas where one is confronted with the geometry-induced, atmosphere-induced, and land-coverage-induced diculties to successfully monitor deformation. The role of nonlinear deformation requires a comprehensive investigation. Typically, interferometric phase models assume a linear deformation. However, mass movements along unstable slopes may exhibit a kinematic behavior that is more complex, spatially variant and influenced by geomorphological factors. Such a behavior would entail improved phase modeling. These multiple lines of research require formal investigation via three doctoral studentships. One of these studentships will be at ITU, Lahore.

Project Partners

The project partners are: Dr. Othmar Frey (ETH Zurich), Dr. Adnan Siddique (ITU, Lahore) & Dr. Farooq Ahmed (UET, Lahore).

Status

Closed successfully in summer 2024.

Summary

Accidental oil spills or dumping by oil tankers in the open waters is a threat to marine and coastal ecosystems. Synthetic-Aperture Radar (SAR) is a useful tool for analysing oil spills, because it operates in all-day, all-weather conditions. An oil spill can typically be seen as a dark stretch in SAR images and can often be detected through visual inspection. The major challenge is to differentiate oil spills from look-alikes, i.e., low-wind areas, algae blooms and grease ice, etc., that have a dark signature similar to that of an oil spill. It has been noted over time that oil spill events in Pakistan’s territorial waters often remain undetected until the oil reaches the coastal regions or it is located by concerned authorities during patrolling. A formal remote sensing-based operational framework for oil spills detection in Pakistan’s Exclusive Economic Zone (EEZ) in the Arabian Sea is urgently needed. In this project, we develop a methodology based on the use of an encoder–decoder-based convolutional neural network trained on an annotated dataset comprising selected oil spill events verified by the European Maritime Safety Agency (EMSA). The dataset encompasses multiple classes, viz., sea surface, oil spill, look-alikes, ships and land. We processed Sentinel-1 acquisitions over the EEZ from January 2017 to December 2023, and we thereby prepared a repository of SAR images for the aforementioned duration. This repository contained images that had been vetted by experts, to trace and confirm oil spills. We tested the repository using the trained model, and, to our surprise, we detected 92 major previously unreported oil spill events within those seven years. In 2020, our model detected 26 oil spills in the EEZ, which corresponds to the highest number of spills detected in a single year; whereas in 2023, our model detected 10 oil spill events. In terms of the total surface area covered by the spills, the worst year was 2021, with a cumulative 395 sq. km covered in oil or an oil-like substance. On the whole, these are alarming figures.

Relevant Publications

1. Basit, A., Siddique, M.A., Bashir, S., Naseer, E. and Sarfraz, M.S., 2024. Deep Learning-Based Detection of Oil Spills in Pakistan’s Exclusive Economic Zone from January 2017 to December 2023. Remote Sensing, 16(13), p.2432.

2. Basit, A., Siddique, M.A., Bhatti, M.K. and Sarfraz, M.S., 2022. Comparison of CNNs and vision transformers-based hybrid models using gradient profile loss for classification of oil spills in SAR images. Remote Sensing, 14(9), p.2085.

3. Basit, A., Siddique, M.A. and Sarfraz, M.S., 2021, July. Deep learning based oil spill classification using Unet convolutional neural network. In 2021 IEEE International Geoscience and Remote Sensing Symposium IGARSS (pp. 3491-3494). IEEE.

We would like to thank Multimedia Knowledge and Social Media Analytics Laboratory (MKLab), ITI-CERTH, Greece, for providing a benchmark dataset for classification of oil spills. This dataset has been used extensively in this project.

Status

The project has closed successfully.

Summary

AI4GLOF will focus on the use of AI for Glacial Lake Outburst Floods hazard potential assessment in Chitral valley, Pakistan. The project aims at bringing in the much-needed tech-intervention to close the gaps between the effects of climate change on glaciers and their subsequent impact on the local communities residing downstream as well as the government’s policies and preparedness to respond in a timely and effective manner. The outcomes of this project will benefit all stakeholders, particularly the local communities, National and International NGOs and the disaster management authorities working in the glaciated regions of Pakistan. The project will provide an integrated analysis of the data from satellite imagery and ground-based weather stations to monitor the evolution of glaciers and assess the potential of hazards due to the formation, evolution and outburst glacial lakes. The project will build national capacity for long-term predictive analysis and early warning systems for disaster assessment & management that will provide critical technical advisory to the policy makers and sufficient time to the authorities to respond to any developing disaster in the glaciated regions of the country.

Project team

The Principle Investigator (PI) of the project is Dr. Adnan Siddique, and Dr. Khurrum Bhatti and Dr. Mohsen Ali are co-PIs. Dr. Brent Minchew and Ms. Tooba Fatima are on board as foreign and local consultants, respectively.

Quantifying the Role of Subglacial Hydrology in Glacier Surges and Outburst Floods of High Mountain Asia