Earth observation for improving forest carbon estimation and degradation monitoring in Bangladesh

About the training

The SERVIR-HKH Initiative and SilvaCarbon are organising a training workshop on improving above-ground forest carbon estimation and degradation monitoring in Bangladesh. The training will cover four components that support forest carbon emission monitoring efforts: terrestrial laser scanning (TLS), global ecosystem dynamics investigation (GEDI), SAR (Synthetic Radar Aperture), and continuous degradation detection (CODED) algorithm.

Participants will learn necessary skills and knowledge on different methods leveraging open-source remote sensing data for improving above-ground forest carbon estimation of Bangladesh. They will be familiar with Online Biomass Inference using Waveforms And iNventory (OBIWAN), GEDI-incorporated model-based above-ground biomass (AGB) estimation, and tools to understand timeseries wall-to-wall biomass estimation. They will also learn TLS data for biomass estimation at the individual tree scale, which will help improve the capacity to scale biomass estimation from the plot to the regional level. Other method is using SAR in improving carbon estimation and wall-to-wall mapping. Participants will also be introduced to CODED algorithms for tracing forest degradation.

We developed this training content and delivery in collaboration with NASA SERVIR Applied Sciences Team (AST), and the Government of Nepal’s Forest Research and Training Centre (FRTC).

Objectives

• Familiarise participants with different methods and data
• Share background and updates on the CODED algorithm, used by FRTC for forest degradation monitoring, and discuss further improvements for the application of CODED
• Seek feedback from participants regarding the different methods and their needs. This feedback will help refine the methods and models to meet the specific needs of stakeholders
• Bring together relevant stakeholders who are actively involved in forest carbon-related works and management. The training workshop will provide a platform for them to share their experiences, challenges, and emerging forest monitoring requirements

Expected outcomes

Upon completion of the training, the participants will have a better understanding of the applications of annual forest cover status in Bangladesh and tools including RLCMS, TLS, GEDI, SAR, and CODED in monitoring forest condition and carbon stocks, enabling them to apply the knowledge in their working areas. At the end of the training, participants will outline a feasible method and data appropriate for the country.

Expected participants

We will organise the in-person training for 20 participants from national agencies in Bangladesh, including Bangladesh Forest Department (BFD), Bangladesh Space Research and Remote Sensing Organization (SPARSO), United States Forest Service, Community Partnerships to Strengthen Sustainable Development Program (USFS COMPASS), universities and other relevant organisations.

Resource persons/facilitators

ICIMOD

Birendra Bajracharya, Chief of Party, SERVIR-HKH

Rajesh Bahadur Thapa, Science and Data Lead, SERVIR-HKH

Kabir Uddin, Remote Sensing Specialist

Sajana Maharjan, Remote Sensing and Geo-information Analyst

BFD

Zaheer Iqbal, Deputy Conservator of Forests (DCF), RIMS Unit

FRTC

Bimal Kumar Acharya, Senior Remote Sensing Officer

Amul Kumar Acharya, Forest Research Officer

U.S Forest Service International Program, Compass

Md. Shams Uddin, Natural Resource Management Lead

U.S Forest Service

Sean Healey, Lead – NASA SERVIR Applied Science Team, and Research Ecologist

Ashraful Islam, Remote Sensing and GIS Associate

Background

This training workshop incorporates five components that support forest carbon emission monitoring efforts.

The first component, regional land cover monitoring system (RLCMS) had generated annual land cover map of HKH region from 2000-2022. It is operational system that updates land cover map annually leveraging open Earth observation data following state-of-the-art methods. This provides opportunity to prepare activity data.

Secondly, SAR microwave imaging system with cloud penetrating and all-day data acquisition capabilities, provide viable alternatives to effectively measure and monitor forest biomass and forest carbon at national scales. Further, there are opportunities to integrate optical and SAR data, which help in improving forest biomass estimation. Next is TLS, which provides comprehensive data on forest structure, enabling more precise biomass estimation compared to traditional methods.

Fourth is GEDI, is a spaceborne LiDAR sampling instrument, collecting waveform return data for 25-m footprints spaced at 60-m intervals along parallel ground tracks that are approximately 600 m apart. Transparent methods have been developed to use GEDI to infer mean biomass with clear estimates of uncertainty. These methods have been built into OBIWAN, an interactive public tool developed to address carbon estimation in individual jurisdictions.

Lastly, the CODED algorithm relies upon time series analysis of the forest component of a spectral unmixing process applied to all available Landsat imagery. It is a leading algorithm for mapping forest degradation, which is canopy disturbance that does not result in deforestation. The amount of area affected by degradation, and its carbon consequences, are often greater than those of forest conversion.