Processing Chandrayaan 2 TMC-2 Images¶

More info on Chandrayaan 2:

Environment¶

You will need:

ALE >= 1.1.2
ISIS >= 10.0.0
SpiceQL >= 1.2.7
usgscsm

To install these with conda:

conda create -n ch2 \
    -c conda-forge -c usgs-astrogeology \
    matplotlib ale=1.1.2 "usgs-astrogeology/label/RC::isis>=10.0.0"
conda activate ch2 # activate env

Warning

As of writing, the latest version of ISIS that supports Chandrayaan-2 is ISIS 10.0.0 RC1, once ISIS 10.0.0 is out of release candidate (RC) status, you can download it with the command:

conda create -n ch2 \
    -c conda-forge -c usgs-astrogeology \
    matplotlib ale=1.1.2 SpiceQL >= 1.2.7 "usgs-astrogeology::isis>=10.0.0_RC1"
conda activate ch2 # activate env

Image Data, Label, and Template¶

The .img image and .xml label are required to import a TMC image into the ISIS cube format. These can be downloaded ISRO's interactive map or ISRO's PRADAN system for bulk downloads. (New users must register an account.)

Spice Kernel Coverage

SPICE kernels are downloaded directly from ISRO's PRADAN system, SPICE coverage isn't complete so you can download images that do not have kernels available for them yet. Either because PRADAN has no kernels for them or we haven't downloaded them yet as it's a manual process.

Check available kernels with the following CURL command:

curl -XGET "https://astrogeology.usgs.gov/apis/spiceql/latest/searchForKernelsets?spiceqlNames=\[chandrayaan2\]&limitCk=-1&limitSpk=-1" | jq

The filenames for CKs and SPKs tell you their time coverage.

Example Data

To follow along you can download an example image from here.

Set Environmental Variables¶

ISIS requires ISISROOT and ISISDATA to be set. You can set ISISROOT equal to the CONDA_PREFIX as below. You will need to set up the ISIS Data Area and set ISISDATA to point to it.

Create a ISIS compatible Cube or GTiff¶

Once set up you can import with isisimport

isisimport from=ch2_tmc_nra_20230214T2032466433_d_img_n18.xml to=ch2_tmc_nra_20230214T2032466433_d_img_n18.cub 
# create and ISD, this creates a file ch2_tmc_nra_20230214T2032466433_d_img_n18.json 
isd_generate -s ch2_tmc_nra_20230214T2032466433_d_img_n18.cub

Local Kernels

isd_generate -s assumes you have local SPICE kernels set up, so make sure you have ISISDATA set with base and chandrayaan2 installed. You can install these data with:

downloadIsisData base $ISISDATA 
downloadIsisData chandrayaan2 $ISISDATA

You can troubleshoot isd_generate with -v, most common issue is missing kernels covering the images time range.

Avoiding ISISDATA Downloads

You can skip ISISDATA downloads using isd_generate -w for web camera generation. This is feature is experimental so use with caution. For example:

isd_generate -w ch2_tmc_nra_20230214T2032466433_d_img_n18.cub

Check you have the files, you will need the .json and .cub

ls -1
# ch2_tmc_nra_20230214T2032466433_d_img_n18.cub
# ch2_tmc_nra_20230214T2032466433_d_img_n18.img
# ch2_tmc_nra_20230214T2032466433_d_img_n18.json
# ch2_tmc_nra_20230214T2032466433_d_img_n18.xml

From here you can use the .cub and .json for any CSM compatible tool like Socet GXP or Ames Stereo Pipeline.

Create an ISIS-compatible image¶

To use this image in ISIS, you will need to combine the JSON ISD with the cube. These can be written as an ISIS Cube or Gtiff. Although, we recommend GTiffs given the large size of CH2 images.

```
mamba install usgscsm # if not installed already 
csminit from=ch2_tmc_nra_20230214T2032466433_d_img_n18.cub isd=ch2_tmc_nra_20230214T2032466433_d_img_n18.json
cubeatt from=ch2_tmc_nra_20230214T2032466433_d_img_n18.cub to=ch2_tmc_nra_20230214T2032466433_d_img_n18.tiff+GTIFF
```

You can view the image with qview, warning, image is very large and might take a minute to render. Gtiffs can be rendered with anything supporting Tiffs, but pixel scaling might cause them to render as black images. We recommend qview, qgis, other tools designed for spatial data.

qview ch2_tmc_nra_20230214T2032466433_d_img_n18.tiff

TMC2 qview

Test image on the left, more typical TMC2 image on the right. Test image is smaller than typical TMC2 image. As you can see, TMC images stretch very far (~1/3 of the planets latitude), image on the left is 74k lines, image on the right is 368k lines. Apps like ISIS's crop can be used to trim the image while maintaining camera data intact.

From here you can use the image in other ISIS apps such as footprintinit and cam2map. Initial pointing in TMC2 has errors (as most instruments). So bundle adjustment is necessary for accuracy.

See for bundle adjustment info: https://astrogeology.usgs.gov/docs/how-to-guides/image-processing/bundle-adjustment-in-isis/

See for stereo info in ASP (similar for TMC2): https://stereopipeline.readthedocs.io/en/latest/examples/chandrayaan2.html#stereo