Urban Tree Canopy Cover
Tier: 3
Data Vintage: 06/2018
Metric Definition and Relevance: This 1m raster layer represents tree canopy cover for the 2010 Census Bureau urban areas for the Northern California RRK region. 
Data Resolution:  1m Raster
Data Units:  Binary, 1 = tree canopy cover
Creation Method:  The tree canopy datasets were processed by Earth Define LLC using a deep convolution neural network-based architecture called SegNet (https://urldefense.com/v3/__https://arxiv.org/abs/1511.00561__;!!Mih3wA!HdYAz62gIB8vz84c7nMYt8f8b_311YDGA3BpgaJBOGrFEe-TBfGNt10ClxdwGgv67rjJaeDCemx_ogoSLCMllQEsdQ$ ). This is an upgrade to the CART (classification and regression trees) based architecture that we previously used for the 2012 tree canopy classification. The deep convolution neural networks (DCNN) based architecture is significantly more robust at handling the radiometric and seasonal variation in the source imagery to provide a more consistent and accurate classification across California.
The tree canopy dataset for California was clipped to the RRK Northern California boundary and kept in ESRI File Geodatabase format. To download original datasets, statewide or by census urban area, go to https://urldefense.com/v3/__https://www.fs.usda.gov/detail/r5/communityforests/?cid=fseprd647385__;!!Mih3wA!HdYAz62gIB8vz84c7nMYt8f8b_311YDGA3BpgaJBOGrFEe-TBfGNt10ClxdwGgv67rjJaeDCemx_ogoSLCMHicEINA$ 
Data Source: 
CALFIRE Fire and Resource Assessment Program | CAL FIRE
Earth Define LLC, EarthDefine | Geospatial Data and Services
California National Agriculture Imagery Program (NAIP) 2018 orthoimagery
Light Detection and Ranging (Lidar)

File Name:  urbancanopy_2018.gdb