Skip to content

Latest commit

 

History

History
57 lines (40 loc) · 3.57 KB

HOWTO.md

File metadata and controls

57 lines (40 loc) · 3.57 KB
Raw

Tutorial: How to run the NPS night-time radiance calculation

Clip_raster_and_calculate_ALR_Python.py takes in a compressed tgz of night-time radiance composites from the VIIRS DNB, processed by the Earth Observations Group. You can download the composites by month and by location here.

The script reads the compressed file into memory, clips to a buffered polygon specified by the user and reprojects to an equal-area Albers projection. It then uses a neighborhood annulus method to calculate the effect of night-time light at varying distances from a source.

Finally, it reclips the result to the region of interest and saves the output to a destination folder.

Because the algorithm calculates the effects of light up to 300 km away from an area of interest, a buffer of 300 km around the user-provided polygon will be generated and used to calculate ALR; the output will later be reclipped to the original polygon. Keep this in mind when selecting tile images from the VIIRS DNB.


An example of how the satellite data is buffered and cropped.

Before running the code

  1. Download the nighttime radiance composites from the VIIRS DNB, available here. Pick the tile that corresponds to your region.
  2. Make sure you have a polygon that represents the region you're interested in.
  3. Install the necessary dependencies (see below).

Set-up

Start by opening Clip_raster_and_calculate_ALR_Python.py and navigate to the ## USER INPUTS section, line 24. You'll want to change all of these variables to match your set-up.

  • working_dir is where you want to store all your outputs.
  • geodb is the location of a File GDB you're using to store the feature class representing the region you're interested in. If you're using a shapefile, set this equal to False.
  • roi is the region of interest, either the name of a feature class or path to a shapefile.
  • name is used for labeling raster output.
  • out_proj should be an equal-area Albers projection, to ensure accuracy with the convolution process. Look up the EPSG code that corresponds to the correct projection for your region.
  • image_tile is the tile value selected from the VIIRS/DNB composites.
  • verbosity gives you the option to log output as the program runs.

You should not need to change any other inputs!

Dependencies:

  • NumPy
  • Astropy
  • Geopandas
  • Gdal

You can run the code via:

python Clip_raster_and_calculate_ALR_Python.py [path_to_DNB_tgz_file.tgz]

The outputs will be saved to a folder called ALR_outputs.

To calculate ALR for multiple months, you can write a batch script such as:

REM cd ..\Documents\Code

for %%I in (..\..\Data\RasterData\DNB_VIIRS_EOG\May_Nov_data\*.tgz) do (echo %%I & python Code_to_calculate_ALR\Clip_raster_and_calculate_ALR.py %%~fI)

Other

  • CreateMosaicDataset.py is a quick and dirty way to create a mosaic dataset of your ALR raster outputs, which can then be manipulated to display changes in ALR over time.
  • Reclassify_ALR_rasters.py is similarly a quick way to produce rasters classified by night-sky quality classes, i.e. good, moderate, poor and Milky Way invisible, per Duriscoe et al.
  • ALR_raster_to_time-aware_polygon.py generates a polygon feature class from the ALR raster outputs, calculates area and percent cover, and adds a date field that can be used to publish time-aware data on ArcGIS Online.