SENTINEL: Difference between revisions

From GRASS-Wiki
Jump to navigation Jump to search
(work in progress 1)
Line 42: Line 42:


==Sentinel-2 in GRASS GIS==
==Sentinel-2 in GRASS GIS==
To exemplify the handling of Sentinel-2 data with GRASS GIS, we will use S2 L1C scene from North Carolina (NC) from November 2016. Here are the details:
S2A_OPER_PRD_MSIL1C_PDMC_20161111T123605_R097_V20161110T160502_20161110T160502
830f2a5a-9914-427c-b4e0-69c6c14c71d4
2016-11-10T15:05:02Z
<!--
S2A_OPER_PRD_MSIL1C_PDMC_20161001T232624_R097_V20161001T160052_20161001T160426
44b7fb4c-d2d7-4739-9b8e-3639659505db
2016-10-01T16:00:52.000Z
-->
In the zipped folder delivered by ESA, the data itself is hidden under:
S2A_OPER_PRD_MSIL1C_PDMC_20161111T123605_R097_V20161110T160502_20161110T160502.SAFE/GRANULE/S2A_OPER_MSI_L1C_TL_MTI__20161110T210803_A007244_T17SPV_N02.04/IMG_DATA/
S2A_OPER_MSI_L1C_TL_MTI__20161110T210803_A007244_T17SPV_B01.jp2
S2A_OPER_MSI_L1C_TL_MTI__20161110T210803_A007244_T17SPV_B02.jp2
S2A_OPER_MSI_L1C_TL_MTI__20161110T210803_A007244_T17SPV_B03.jp2
S2A_OPER_MSI_L1C_TL_MTI__20161110T210803_A007244_T17SPV_B04.jp2
S2A_OPER_MSI_L1C_TL_MTI__20161110T210803_A007244_T17SPV_B05.jp2
S2A_OPER_MSI_L1C_TL_MTI__20161110T210803_A007244_T17SPV_B06.jp2
S2A_OPER_MSI_L1C_TL_MTI__20161110T210803_A007244_T17SPV_B07.jp2
S2A_OPER_MSI_L1C_TL_MTI__20161110T210803_A007244_T17SPV_B08.jp2
S2A_OPER_MSI_L1C_TL_MTI__20161110T210803_A007244_T17SPV_B09.jp2
S2A_OPER_MSI_L1C_TL_MTI__20161110T210803_A007244_T17SPV_B10.jp2
S2A_OPER_MSI_L1C_TL_MTI__20161110T210803_A007244_T17SPV_B11.jp2
S2A_OPER_MSI_L1C_TL_MTI__20161110T210803_A007244_T17SPV_B12.jp2
S2A_OPER_MSI_L1C_TL_MTI__20161110T210803_A007244_T17SPV_B8A.jp2
<!-- explain naming -->


===Import data===
===Import data===


Sentinel-2 data comes in JPEG2000 format that can be easily imported into GRASS GIS either by means of {{cmd|r.in.gdal}} (in a UTM location) or using {{cmd|r.import}} that provides reprojection on the fly.  
Sentinel-2 data comes in JPEG2000 format that can be easily imported into GRASS GIS either by means of {{cmd|r.in.gdal}} (in a UTM location) or using {{cmd|r.import}} that provides reprojection on the fly. In this example, we will use {{cmd|r.import}} to reproject and import into the NC sample location delivered with GRASS GIS.  


Example here
<source lang="bash">
for map in `ls *.jp2` ; do
outmap=`basename $map .jp2`
r.import input=$map output=$outmap
done
</source>


===Atmospheric correction with {{cmd|i.atcorr}}===
===Atmospheric correction with {{cmd|i.atcorr}}===
Line 53: Line 88:
Example here
Example here


<!--
===Apply cloud mask===
===Apply cloud mask===
Example here
Example here
-->


===Estimate diverse indices===
===Estimate diverse indices===

Revision as of 16:23, 24 September 2017

Sentinel 2

Sentinel-2 is an Earth observation mission developed by ESA as part of the Copernicus Programme to perform terrestrial observations in support of services such as forest monitoring, land cover changes detection, and natural disaster management. It consists of two identical satellites, Sentinel-2A and Sentinel-2B. Details about the mission, user and technical guides as well as other documentation can be found at the ESA dedicated website [1] and a number of other sites.

Overview of Sentinel-2 mission capabilities

Sentinel-2 mission has the following features:

  • Multi-spectral data with 13 bands in the visible, near infrared (NIR), and short wave infrared (SWIR).
  • Systematic global coverage of land surfaces from 56° S to 84° N, coastal waters, and all of the Mediterranean Sea
  • Revisiting every 5 days under the same viewing angles. At high latitudes, Sentinel-2 swath overlaps and some regions are observed twice or more every 5 days, but with different viewing angles.
  • Spatial resolutions of 10 m (Bands 2, 3, 4 and 8 - Blue, Green, Red and NIR), 20 m (Bands 5, 6 and 7 - Red Edge bands, Band 8A - Narrow infrared and Bands 11 and 12 - SWIR) and 60 m (Bands 1 - Coastal aerosol, Band 9 - Water vapour and Band 10 - SWIR/Cirrus)
  • 290 km field of view
  • Free and open data policy

Sentinel-2 products

  • Level-1B product provides radiometrically corrected imagery in Top-Of-Atmosphere (TOA) radiance values and in sensor geometry. Level-1B product is composed of an ensemble of granules that are 25 km across track (AC) by 23 km along track (AL). Additionally, this product includes the refined geometry which is used to generate the Level-1C product. Level-1B pixel coordinates refer to the centre of each pixel.
  • Level-1C product is composed of 100 km² single tiles (ortho-images in UTM/WGS84 projection). The Level-1C product results from using a Digital Elevation Model (DEM) to project the image in cartographic coordinates. Per-pixel radiometric measurements are provided in TOA reflectances with all parameters to transform them into radiances. Level-1C products are resampled with a constant Ground Sampling Distance (GSD) of 10, 20 and 60 m depending on the native resolution of the different spectral bands. In Level-1C products, pixel coordinates refer to the upper left corner of the pixel. Before September 2016, Level-1C was delivered as multi-tile product.
  • Level-2A product provides Bottom Of Atmosphere (BOA) reflectance images derived from the associated Level-1C products. Therefore, each Level-2A product is also a 100 km² tile in cartographic geometry (UTM/WGS84 projection). BOA reflectances in Level-2A products are scaled by a factor of 10000.

Data download

Sentinel-2 data might be obtained from the Copernicus Open Access Hub, either by using the online interactive interface and browsing to your area of interest or by means of the API Hub. There are several other sites from which it is possible to download Sentinel-2 imagery as well. Here, some of them:

For those preferring or needing to download data via the API, there also exist several options:

  • sentinelsat: Python utilities to access the API of Copernicus Sentinels Scientific Data Hub (useful for all Sentinel satellite imagery)
  • Sentinel-download: Automated download of Sentinel-2 L1C data from ESA (through wget)

Data pre-processing

As from May 2017, ESA started to provide Level 2A products for Europe and they will reprocess the archive during this year (2017). However, it is (still) possible to download Level-1C products and customize the atmospheric correction by using different algorithms (i.e.: 6S, 6SV, arcsi) and/or a different DEM (Sen2Cor uses STRM 90m by default).

Sentinel-2 in GRASS GIS

To exemplify the handling of Sentinel-2 data with GRASS GIS, we will use S2 L1C scene from North Carolina (NC) from November 2016. Here are the details:

S2A_OPER_PRD_MSIL1C_PDMC_20161111T123605_R097_V20161110T160502_20161110T160502
830f2a5a-9914-427c-b4e0-69c6c14c71d4
2016-11-10T15:05:02Z

In the zipped folder delivered by ESA, the data itself is hidden under:

S2A_OPER_PRD_MSIL1C_PDMC_20161111T123605_R097_V20161110T160502_20161110T160502.SAFE/GRANULE/S2A_OPER_MSI_L1C_TL_MTI__20161110T210803_A007244_T17SPV_N02.04/IMG_DATA/
S2A_OPER_MSI_L1C_TL_MTI__20161110T210803_A007244_T17SPV_B01.jp2
S2A_OPER_MSI_L1C_TL_MTI__20161110T210803_A007244_T17SPV_B02.jp2
S2A_OPER_MSI_L1C_TL_MTI__20161110T210803_A007244_T17SPV_B03.jp2
S2A_OPER_MSI_L1C_TL_MTI__20161110T210803_A007244_T17SPV_B04.jp2
S2A_OPER_MSI_L1C_TL_MTI__20161110T210803_A007244_T17SPV_B05.jp2
S2A_OPER_MSI_L1C_TL_MTI__20161110T210803_A007244_T17SPV_B06.jp2
S2A_OPER_MSI_L1C_TL_MTI__20161110T210803_A007244_T17SPV_B07.jp2
S2A_OPER_MSI_L1C_TL_MTI__20161110T210803_A007244_T17SPV_B08.jp2
S2A_OPER_MSI_L1C_TL_MTI__20161110T210803_A007244_T17SPV_B09.jp2
S2A_OPER_MSI_L1C_TL_MTI__20161110T210803_A007244_T17SPV_B10.jp2
S2A_OPER_MSI_L1C_TL_MTI__20161110T210803_A007244_T17SPV_B11.jp2
S2A_OPER_MSI_L1C_TL_MTI__20161110T210803_A007244_T17SPV_B12.jp2
S2A_OPER_MSI_L1C_TL_MTI__20161110T210803_A007244_T17SPV_B8A.jp2


Import data

Sentinel-2 data comes in JPEG2000 format that can be easily imported into GRASS GIS either by means of r.in.gdal (in a UTM location) or using r.import that provides reprojection on the fly. In this example, we will use r.import to reproject and import into the NC sample location delivered with GRASS GIS.

for map in `ls *.jp2` ; do
 outmap=`basename $map .jp2`
 r.import input=$map output=$outmap
done

Atmospheric correction with i.atcorr

Example here


Estimate diverse indices

Given that Sentinel-2 MSI comes with 3 red edge bands, it allows to estimate several other vegetation indices, aside from the most classic ones. There is a quite exhaustive list at the Sentinel Hub website.

  • EVI = 2.5*(B08 - B04) / (B08 + 6*B04 - 7.5*B02 + 1)
  • ARVI = (B08 - (2*B04 - B02)) / (B08 + (2*B04 - B02))
  • NDII = (B08 - B11) / (B08 + B11)
  • MSI = B11 / B08
  • PSRI-NIR = (B04 - B02) / B08
  • NDVI-GREEN = B03 * (B08 - B04) / (B08 + B04)
  • MTCI = (B06 - B05) / (B05 - B04)
  • IRECI = (B07 - B04) * B06 / B05
  • GRVI1 = (B04 - B03) / (B04 + B03)
  • GNDVI = (B08 - B03) / (B08 + B03)
  • EVI2 = 2.5 * (B08 - B04) / (B08 + 2.4 * B04 + 1)
  • ChlRE = B05 / B08

Classification

Example here

Further reading

IMAGICO.DE blog