Internal scripts of WWF project (deforestration monitoring at Russia Far East)
1x1 degree grid, 3 TIFF files for each cell (all TIFFs have single int16 channel):
r_date_c.tif- data, number of day from the beginning of the year (1-365)r_loss_c.tif- loss probability, 0-100cong_c.tif- loss confidence (0-3)
Original files are processed by 16bit2rgb.exe (not included) to make single 8-bit RGB TIFF for each cell. Node.js script combine_16bit2rgb.js is used to process all the cells.
Format of the RGB file is the following:
- R channel: low 8 bits of date
- G channel
- bits 0-3: bits 8-11 of date
- bits 6-7: confidence
- B channel: loss probability
RGB composites are processed by GeoMixer Tiling Tool (see run_tiling.bat) and uploaded to GeoMixer server as raster layer.
frontend/Leaflet.WWFAlarm.js browser Leaflet plugin is used to show that raster layer on a map (only API, no UI).
frontend/WWFPlugin.js is a plugin for GeoMixer Editor to show this layer in Editor.
Rest python and Node.js scripts are used to vectorize RGB composites:
process.py(combine.js) - extract clusters separately by datesprocess_single.py(combine_single.js) - extract clusters without taking dates into consideration
Instantiation example
var alarm = new L.WWFAlarm('CAABA6E0E4FF45319C731E56B1064D0D', '1D116AC56D694C0B8CDCA87C06D1961A');
Methods:
setDateInterval(daysMin, daysMax): show changes only within given date interval (daysMin, daysMax - number of days from January 1st)setLossThreshold(minLoss): minLoss from 0 to 100setConfThreshold(minConf): minConf from 0 to 3