the redd+ satellite based land cover monitoring system for mexico
DESCRIPTION
Remote sensing –Beyond images Mexico 14-15 December 2013 The workshop was organized by CIMMYT Global Conservation Agriculture Program (GCAP) and funded by the Bill & Melinda Gates Foundation (BMGF), the Mexican Secretariat of Agriculture, Livestock, Rural Development, Fisheries and Food (SAGARPA), the International Maize and Wheat Improvement Center (CIMMYT), CGIAR Research Program on Maize, the Cereal System Initiative for South Asia (CSISA) and the Sustainable Modernization of the Traditional Agriculture (MasAgro)TRANSCRIPT
The REDD+ satellite based
land cover monitoring
system for Mexico
CIMMYT Remote Sensing Workshop, 14./15.12.2013, Mexico City, MexicoSteffen Gebhardt, CONABIO, [email protected]
Objectives
Activity Data (AD) monitoring within REDD+ is primarily based on wall-to-wall land cover and land cover change information.
Automatic satellite image classification is required to assure timely product generation in a standardized and cost-beneficial manner especially for a
country the size of Mexico.
Operative satellite forest monitoring system implemented by CONABIO within the Mexican-Norwegian Project Reinforcing REDD+ Readiness in Mexico and
enabling South-South cooperation.
REDD+ Measuring, Reporting and Verification (MRV) system
System Specifications
MRV system elements
IPCC elements
ContextEmission and removals from forests
IPCC basic method
Activity Dataland representation
Satellite Forest Monitoring system
Emission Factors C stock changes
National Forest Inventory
Emission estimatesGHG emissions and removals
National GHGs Inventory
X =
Operational wall-to-wall system based on satellite remote sensing data, with a sampling approach to assess historical deforestation and degradation rates. Changes in forest area to be assessed in order to fulfil the IPCC Tier 3 reporting requirements
INFyS implemented in 2004. Consistent and comparable over time, revision in 5 year interval. Data on carbon stock for all forest carbon pools for the main forest types at IPCC Tier 2 and Tier 3 reporting requirements.
National inventory for the LULUCF sector developed following the reporting requirements of the Annex-I Parties under the UNFCCC. Following the IPCC default methods: ‘gain-loss’ or ‘stock difference’, but it could also be developed to implement a Tier 3 model.
Operational Processing System
• “The Measuring, Reporting and Verification - Activity Data (MRV-AD) Monitoring System within the Mexican REDD+ program” =
MAD-Mex
• Products at 1:100,000 and 1:20,000
• Land Cover (LC), Land Cover Change (LCC)
• Forest / Non-Forest, Forest Change (FC)
• Cover density
• Automatic classification by MAD-Mex and subsequent visual interpretation to 60 classes in agreement with INEGI
• Base Line starting 1990-2020 (Landsat 5,7,8) and operational yearly monitoring 2011-2020 (RapidEye)
Operational Processing System
Processing
Storage
Workflows / Processes
Software
MAD-Mex
Remote Sensing Data for AD Monitoring
Landsat 135 distinct tiles
Remote Sensing Data for AD Monitoring
RapidEye 4000 distinct tiles
MAD-Mex Landsat LCC method
MAD-Mex Landsat LCC products
MAD-Mex Landsat LCC products
MAD-Mex Landsat LCC accuracies
Run 1 Run 2 Run 3 Run 4 Run 5
Temperate forest 82.1 80.5 79.3 81.2 78.8
Tropical forest 77.3 76.9 76.2 77.5 77.0
Scrubland 80.7 80.7 80.7 80.7 80.7
Wetland vegetation 66.7 64.8 66.7 64.8 68.5
Agriculture 77.0 76.9 75.4 78.5 76.0
Grassland 62.2 61.6 62.2 62.2 62.5
Water body 68.9 66.2 59.5 64.9 64.9
Barren land 72.0 88.0 80.0 80.0 84.0
Urban area 67.2 73.4 67.2 67.2 64.1
1993 76.2 75.8 76.1 76.1 76.1
1995 75.7 75.7 76.3 77.1 76.7
2000 74.8 76.2 75.7 75.8 75.3
MAD-Mex RapidEye LCC method
Escalas 1:250,000 vs. 1:100,000 vs. 1:20,000
MAD-Mex Landsat vs. RapidEye
MAD-Mex Landsat vs. RapidEye
MAD-Mex Landsat vs. RapidEye
MAD-Mex Landsat vs. RapidEye
MAD-Mex Landsat vs. RapidEye
MAD-Mex Landsat vs. RapidEye
MAD-Mex Landsat vs. RapidEye
MAD-Mex Landsat vs. RapidEye
MAD-Mex Landsat vs. RapidEye
MAD-Mex RapidEye Change Detection
Change Intensities
Strong negative
Medium negative
Light negative
No change
Light positive
Medium positive
Strong positive
2010-01-24 2011-03-10
MAD-Mex RapidEye Change Detection
MAD-Mex RapidEye Change Detection
MAD-Mex RapidEye Change Detection
MAD-Mex RapidEye Change Detection
Highlights
• The Measuring, Reporting and Verification - Activity Data (MRV-AD) Monitoring System within the Mexican REDD+ program (MAD-Mex) enables automatic wall-to-wall land cover classification.
• Using Landsat data seven national land cover maps at a scale of 1:100,000 between 1993 and 2008 have been generated yielding in overall accuracies up to 76% over 9 land cover classes. Tropical and temperate forest was classified with accuracy up to 78% and 82%, respectively.
• A first and preliminary national land cover product at a scale of 1:20,000 using RapidEye data of 2011 is expected by the end of the year.
• Thank you
• steffen.gebhardt rainer.ressl michael.schmidt @conabio.gob.mx