Error as occured...

Side Event #6671

Status: SCHEDULED

Date Thursday
2024.10.31 @ 13:20
Add to Calendar
Room Paramos - NGOs meeting room
Plaza One
Capacity: 135 people
Conference SBI-05 / COP-16 / CP-MOP-11 / NP-MOP-05
Meetings(s)
COP-16
Sixteenth meeting of the Conference of the Parties to the Convention on Biological Diversity
Title Enhancing the Assessment of Mining Impacts on Biodiversity through Artificial Intelligence and Earth Observation Data: A Novel Approach for Sustainable Detection and Management of Mining Sites
Hosts
UNCCD | ANDI |
 

 

1. Introduction and Rationale:

Approximately 11.5% of the world's land has been affected by mineral extraction activities. After mining operations end, these abandoned sites often have low levels of organic carbon and minerals, which impedes the growth of colonizing species. In South Africa, mining is a significant economic driver, contributing around 8% of the GDP and providing employment to a large segment of the workforce. The sector is crucial for the country’s economic stability and growth. However, mining operations in South Africa frequently result in severe land degradation, such as deforestation, soil erosion, and biodiversity loss. They also pose risks to water quality through the introduction of heavy metals and other pollutants, with acid mine drainage being a particularly severe issue as sulfuric acid leaks into water from exposed mineral deposits. Moreover, mining produces large amounts of waste rock and tailings, which can lead to significant environmental problems if not managed properly. In Colombia, official data shows that 42.1 tons of gold were produced in 2022, but it is estimated that only 13% of this was mined legally. Illegal mining in Colombia is associated with serious issues including deforestation in protected areas, contamination of water bodies and soil with toxic waste, and the financing of illegal groups.

Given the rising demand for mineral resources and the need for sustainable management, it is vital to develop effective geospatial methods for monitoring mining activities. This study presents a novel decision-making model that employs various spectral indices to detect and achieve sustainable monitoring of mining operations. Detecting mining activities from the ground can be difficult, especially when analyzing extensive and complex areas. This study explores the use of Earth observation data and advanced technologies—particularly Deep Learning and multispectral satellite imagery—to analyse and monitor mining sites and tailings dams in South Africa and Colombia. The aim is to evaluate their status over time and identify areas in need of land restoration.

This study demonstrates the effectiveness of using Artificial Intelligence, specifically Deep Learning techniques, to identify mines and tailings dams with freely available multispectral satellite images: from Sentinel-2 and Landsat 8, combined with MODIS data, leveraging open-source software and reusable code such as AI (deep learning), cloud computing through Google Earth Engine, and various software and libraries (TensorFlow, Google Colaboratory, QGIS, Python, and R).

The side event  aims to showcase the effectiveness of combining Artificial Intelligence, particularly Deep Learning techniques, with Earth observation data to identify mines and tailings dams using freely available multispectral satellite images. This advancement, introduced by the G20 Global Land Initiative at the UNCCD, represents a significant leap forward in monitoring mining activities. It enables the tracking of these activities over time and the assessment of their operational status across extensive areas, with potential applications on a global scale.

2. Key Themes and Discussion Points:

  1. Integration of AI and Earth Observation:
  • Application of Deep Learning Techniques: Exploring how AI, especially Deep Learning, improves the identification and monitoring of mining activities.
  • Integration of AI with Satellite Data: Benefits of combining multispectral satellite images with AI to enhance accuracy and efficiency in monitoring mining activities
  1. Case Studies and Real-World Examples:
  • Examples from Colombia and South Africa: Insights from case studies demonstrating the practical application of these technologies.
  • Lessons Learned: Key takeaways and challenges encountered in implementing AI-driven monitoring systems for biodiversity conservation

 

3. Targeted audience:

  • Environmental and Conservation Organizations
  • Government Agencies and Policy Makers
  • Academia and Research Institutions
  • Technology and AI Specialists:
  • Mining Industry Professionals
  • Non-Governmental Organizations (NGOs

4. Potential Speakers (up to 3 invitees per session)

The side event speakers will discuss the challenges and lessons learned from successful mining restoration projects. They will provide insights into the impacts of large-scale mining and demonstrate how modern technologies can be used to detect mining activities and identify restoration opportunities.

  1. Dr. Orlando Rangel: National University of Colombia, Faculty of Sciences, Institute of Natural Sciences: Expert in fauna and flora of Colombia.
  2. Dr. Hamlet Valois: Research Group on Ecology and Conservation of Tropical Ecosystems, Faculty of Natural Sciences, Chocó Technological University Diego Luis Córdoba: Expert in community restoration projects in illegal mining areas in Chocó – Colombia.
  3. Dr. Luis Madriñán: Environmental manager at Cerrejón. Cerrejón is an open-cast coal mining company operating in northern Colombia. It has restored more than 4,700 hectares.
  4. Dr. Mohamed Abd salam El Vilaly, Programme Officer,G20 Global Initiative Coordination Office, United Nations Convention to Combat Desertification
  5. Luisa Benavides: G20 Global Initiative Coordination Office, United Nations Convention to Combat Desertification

Error as occured...
File Title & Description Tags


 

Error as occured...
File Title & Description Tags


 

Error as occured...
File Title & Description Tags


 

Documents