Monthly global VIIRS mosaic masked with DMSP/OLS

Project name: "Estimating Seasonal Population via Night-Time Satellite Images" (SEPT).

A project funded by Greek General secretariat and Technology & Hellenic Foundation For Research and Innovation.

Spatial Analysis, GIS and Thematic Mapping Laboratory Department of Planning and Regional Development University of Thessaly, School of Engineering

The scripts included in the current repository create global mosaic for suomi npp VIIRS ("vcmsl" version, stray-light corrected data) monthly geotiffs [1] which are masked with values ​​(DN>=6) of 2013 DMSP/OLS data. The goal of the mask application is to exclude ephemeral lights and noise.

The final files are produced in geotiff format, compressed with LZW compression algorithm and saved as unsigned 16-bit integer (UInt16) data type. The mothly mosaics were created for the years 2014, 2015, 2016.

Intercalibrated DMSP/OLS data

The intercalibrated version of DMSP/OLS data [2],[3] were used for the masking of VIIRS tiles. These data are available in the ols_tiles folder. They have been reprojected in WGS84, have been fragmented into tiles corresponding to those of VIIRS and have been converted into binary formats so as DN>=6 => 1 and DN<6 => Null.

Prerequisites

For the successful execution of the process via bash scripts, the following software/libraries must be preinstalled:

• wget, necessary to download the data from the relevant site.
• tar, pv, to decompress and monitor the progress of decompression of the *.tgz files.
• gdalbuildvrt, gdal_translate, to convert the data type of geotiff files and create the global mosaic.
• gdal_calc, to apply the "mask" to the data.

Usage

To complete the process of creating the global mosaic, the following sscripts were executed successively:

• downloadData.sh: downloads the necessary tiles stored in tgz compressed format from the NGDC Earth Observation Group (EOG). Files for download are provided through a file with a list of the URLs of * .tgz files.

• extractData.sh: decompresses the files .avg_rade9h.tif from the compressed *.tgz files.

• float2int.sh: convert geotiffs to UInt16 data type.

• maskData.sh: masks the VIIRS data based on 2013 DMSP/OLS data. Excludes the pixels that they have DN<6 in 2013 DMSP/OLS data.

• mosaic.sh: creates a global mosaic of the VIIRS individual tiles.

• compressData.sh: Compress global VIIRS geotiffs to *tar.gz files .

Example

Monthly VIIRS tiles are available for download in the files 2014/2014.txt, 2015/2015.txt, 2016/2016.txt. An example is given for 2014. To generate the mosaic for 2014 the following commands are required:

• downloadData.sh ./2014/2014.txt ./2014
• extractData.sh ./2014
• float2int.sh ./2014
• maskData.sh ./2014 ./ols_tiles
• mosaic.sh ./2014
• compressData.sh ./2014

Result

The output for 2014, 2015, 2016 is available in osf.io repository.

References

1. NOAA-National Geophysical Data Center, Version 1 VIIRS Day/Night Band Nighttime Lights, Earth Observation Group, NOAA/NCEI. URL https://ngdc.noaa.gov/eog/dmsp/downloadV4composites.html (accessed 05.05.2019).

2. Stathakis, D., 2016. Intercalibration of DMSP/OLS by Parallel Regressions. IEEE Geosci. Remote Sens. Lett. 13, 1420–1424. https://doi.org/10.1109/LGRS.2016.2582784

3. NOAA-National Geophysical Data Center, Defense Meteorological Satellite Program (DMSP), Global DMSP-OLS Nighttime Lights Time Series 1992 - 2013 (Version 4). Earth Observation Group - Defense Meteorological Satellite Progam, Boulder | ngdc.noaa.gov. URL https://ngdc.noaa.gov/eog/dmsp/downloadV4composites.html.

License

This project is licensed under the MIT License - see the LICENSE file for details