Long-term multi-sensor record of fire disturbances in High Northern Latitudes

Tatiana Loboda, Univ. Maryland, loboda@umd.edu (Presenter)
Nancy HF French, Michigan Tech Research Institute (MTRI), nhfrench@mtu.edu
Liza Jenkins, Michigan Tech Research Institute (MTRI), liza.jenkins@mtu.edu
Mary Ellen Miller, Michigan Tech Research Institute, marymill@mtu.edu
Laura Louise Bourgeau-Chavez, Michigan Tech Research Institute, lchavez@mtu.edu

We present a consistent data set in support of the planned NASA Arctic-Boreal Vulnerability Experiment (ABoVE) campaign. This archive of fire observations includes datasets developed from coarse and moderate resolution instruments focused on mapping and characterization of fire events. Here we present a full circumpolar record of fire occurrence 2001 – 2015 above 60 N from the MODIS imagery using a regionally-adapted algorithm. The same algorithm is also adapted to the VIIRS data to assess and ensure burned area product continuity between the sensors. We further evaluate existing MODIS and VIIRS active fire records in the HNL region to provide a quantitative assessment of the compatibility of the long-term record from the two sensors. Our moderate resolution suite includes applications of a new algorithm aimed at detecting and mapping of small fires from Landsat imagery that do not currently appear in the existing databases. Mapping of these small fires, mostly located in the tundra region of HNL, will improve the completeness of the currently available data record and allow for a more robust characterization of fire occurrence in the region. In addition to improving the record of fire occurrence, we use optical (Landsat) and microwave (SAR) data to characterize wildfire impacts within burns using a set of metrics including delta Normalized Burn Ratio (dNBR) and relativized dNBR, both related to state of above ground vegetation; Tasseled Cap Brightness related to removal of above ground vegetation and exposure of soil; and backscatter intensity from SAR which is related to post-fire soil moisture levels. Finally, we apply a Fire Progression Mapping algorithm to infer date of burning within the extent of burned area from MODIS active fire detections. These products will become publicly available within the NASA ABoVE Science Cloud by the beginning of the planned ABoVE campaign in August of 2015.

Presentation Type:  Poster

Session:  General Contributions   (Tue 4:35 PM)

Associated Project(s): 

• Loboda, Tatiana: Social drivers of land cover change around African transboundary Peace Parks ...details

Poster Location ID: 183