ReCover REDD and sustainable forest management EU ReCover project: Remote sensing services to support REDD and sustainable forest management in Fiji Pacific Island GIS&RS conference 2012, 27 – 30 November 2012, Suva Johannes Reiche, Martin Herold: Wageningen University Donata Pedrazzani: GMV Fabian Enßle: Freiburg University
23
Embed
Pacific Island GIS&RS conference 2012, 27 – 30 November 2012, Suva
EU ReCover project: Remote sensing services to support REDD and sustainable forest management in Fiji. Pacific Island GIS&RS conference 2012, 27 – 30 November 2012, Suva. Johannes Reiche, Martin Herold: Wageningen University Donata Pedrazzani: GMV Fabian Enßle: Freiburg University. Outline. - PowerPoint PPT Presentation
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
ReCover for REDD and sustainable forest management
EU ReCover project: Remote sensing services to support REDD and sustainable forest management in Fiji
Pacific Island GIS&RS conference 2012,27 – 30 November 2012, Suva
Johannes Reiche, Martin Herold: Wageningen UniversityDonata Pedrazzani: GMV
Fabian Enßle: Freiburg University
ReCover for REDD and sustainable forest management
Outline
1. ReCover project objective
2. ALOS PALSAR change detection and time-series analysis
3. MODIS time-series analysis for forest change detection
4. ICESat/GLAS space borne laser ranging for forest height & biomass
5. ReCover workshop and field work (October 2012)
2
ReCover for REDD and sustainable forest management
3
1. EU ReCover project objective
• To develop beyond state-of-the-art service capabilities to support reducing deforestation and forest degradation in the tropical regions:– Research project driven by REDD+ monitoring needs– Monitoring system of forest cover, forest cover
changes and biomass mapping including accuracy assessment.
– Capabilities are based on utilizing earth observation and in-situ data
– Using multiple remote sensing data sources– Involvement of national and regional partners, and
user organizations
ReCover for REDD and sustainable forest management
2. ALOS PALSAR change detection and time-series analysis
4
• ALOS PALSAR– L-band SAR system (sensitive to biomass)– SAR is not affected by clouds– Fine Beam Dual data was ordered and processed to 25 m resolution
• Country-wide mosaic for 2010 (25 m) (will be completed)
False colour image RGBR: HH polarisationG: HV polarisationB: HH/HV ratio
ReCover for REDD and sustainable forest management
ALOS PALSAR: Dual-temporal (2007,2010) coverage of west Viti Levu
5
2007-08/092010-08/09
2. ALOS PALSAR change detection and time-series analysis
ReCover for REDD and sustainable forest management
6
ClassificationStep 1: water mask (HH-07&10)Step 2: Vegetation cover change (HV difference 2007-2010)Step 3: Differentiating deforestation and other vegetation decrease, such as agriculture (HH-HV difference 2007)
Water mask
Positive change (e.g. reforestation)
Negative change
Forest/dense vegetation -> non-forest
Other vegetation decrease
Forest land cover change detection (Viti Levu west) 2007 - 2010 (first results, need to be evaluated)
6
Water mask
Positive change (e.g. reforestation)
Negative change
Forest/dense vegetation -> non-forest
Other vegetation decrease
ReCover for REDD and sustainable forest management
Time-series examples
7
Stable forest
2. ALOS PALSAR change detection and time-series analysis
ReCover for REDD and sustainable forest management
8
Deforestation of pine plantagenTime-series examples
2. ALOS PALSAR change detection and time-series analysis
ReCover for REDD and sustainable forest management
9
RegrowthTime-series examples
2. ALOS PALSAR change detection and time-series analysis
ReCover for REDD and sustainable forest management
• BFAST: – time-series analysis package that detects changes as breaks in the time-series – Developed by Dr. Jan Verbesselt, Wageningen University (Netherlands)– BFAST R package is open source and free of charge ('http://bfast.r-forge.r-project.org/)
3. MODIS NDVI time-series for forest change detection using BFAST algorithm (Verbesselt et al.)