Kellndorfer, Josef: Earth Big Data, LLC (Project Lead)
Houghton, Richard (Skee): Woodwell Climate Research Center (Co-Investigator)
Olofsson, Pontus: NASA MSFC (Co-Investigator)
Woodcock, Curtis: Boston University (Co-Investigator)
Cartus, Oliver: Woodwell Climate Research Center (Participant)
Project Funding:
2013 - 2016
NRA: 2013 NASA: Carbon Monitoring System
Funded by NASA
Abstract:
We propose to support the development and improvement of national MRV systems for REDD+ through two objectives. First, we will develop, test, and share with the public domain robust and transparent methods for mapping activity data (e.g., deforestation, forest degradation). Second, we will conduct an uncertainty analysis of carbon emission estimates from the activity data and from emission factors. We will use novel approaches to time series data mining of optical and radar satellite imagery and conduct the work in three test sites (Colombia, Peru and Mexico) identified as National Demonstrator sites by the Group on Earth Observation's (GEO) Forest Carbon Tracking Task (GEO-FCT). The test sites include a variety of ecosystems, biomass regimes, and cloud-cover conditions, and they exhibit a range of drivers of deforestation and land conversion methods, including selective logging, burning, clearing for permanent conversion, and forest regrowth. A large amount of data from optical and radar satellites has already been collected for these GEO-FCT verification sites.
More specifically, we will develop an algorithm from optical and radar time series fusion to produce an accurate assessment of annual changes in areas experiencing deforestation, forest degradation, and forest regrowth (i.e., activity data). The work will include an approach for distinguishing between natural disturbances and permanent anthropogenic change. We will assess the uncertainty and accuracy of the activity data estimated with this algorithm.
To assess the uncertainty of carbon emission estimates, we propose to compile a database of country specific emission factors, stratified by land-cover categories (from the first objective), and linked with carbon density estimates from forest inventory and existing biomass maps. The database will contain uncertainty estimates. To provide guidance for national MRV implementation, we will also explore the impact of uncertainties in activity data and emissions factors on carbon fluxes estimated using a bookkeeping model.
The proposed work is relevant to the specific objectives of this NASA Carbon Monitoring System solicitation, including rigorous exploitation of NASA and international partner satellite remote sensing resources and computational capabilities. The Subsidiary Body of Scientific and Technological Advice (SBSTA) of the UNFCCC agreed in June 2013 that continuous improvement of data and methods is vital for developing MRV systems for REDD+. In particular, SBSTA identified the need to reduce uncertainties in emissions accounting and to develop methodologically consistent ways to harness new observational data, whether field or remote sensing, that can be used to report against reference levels of deforestation and forest degradation, as well as associated reference emission levels (SBSTA, 2013). To develop methodologically consistent, transparent, yet flexible accounting methods, as required in the international framework of the UNFCCC, as well as numerous bi- and multi-lateral agreements, the
Group on Earth Observation (GEO) has established a Forest Carbon Tracking Task (GEO- FCT). PI Kellndorfer and Co-I's Woodcock and Olofsson are among those chosen by GEO to formulate and support the implementation of a Global Forest Observing Initiative (http://geo- fct.org). Support of this proposal would allow them to carry out that work.
More details may be found in the following project profile(s):
