Turbidity_Animated_Timeseries_notebook #1073
Conversation
Proposed real world example notebook "Turbidity_animated_timeseries" created by Joram Downes
Joram95
requested review from
robbibt,
BexDunn,
erialC-P,
uchchwhash,
Matt-dea,
Kooie-cate,
geoscience-aman,
JM-GA,
treefern,
margaretharrison,
vnewey,
Ariana-B,
amanda2099,
supermarkion and
erin-telfer
as code owners
|
Check out this pull request on See visual diffs & provide feedback on Jupyter Notebooks. Powered by ReviewNB |
|
Hey Joram! |
|
Thanks Claire, |
I have suggested a couple of changes and implemented them in this commit. They include: - land masking each image against itself rather than against a single image in the time series. I thought that this would make the outputs more dynamic and will handle tide and flooded inland waterways. - simplified your cloud masking function but its outputs should be unchanged - I have made a separate PR to the dea_tools module to include an NDTI index. (As there was already a tillage index, our version is called NDTI2). I have updated the notebook to use this calculation. - The dea_tools PR also includes an update to the rgb plotting function to enable the addition of titles to subplots. In this case, the titles now note the image timesteps that users can use to identify timesteps for the animations. - I have consolidated the animation timestep selection into a single new variable. - I also added a direct comparison of the animations datasets to confirm that their datestamps are the same. - Finally, I made a couple minor grammatical changes.
I temporarily added the updated script while reviewing the notebook. I'm deleting it now to avoid conflicts with the dea_tools PR that will add these changes to the repo.
I temporarily added the updated script while reviewing the notebook. I'm deleting it now to avoid conflicts with the dea_tools PR that will add these changes to the repo.
There was a problem hiding this comment.
Hey Joram,
Thanks for your hard work in adding this excellent notebook to the DEA-notebooks repo!
I made a small series of changes that are incorporated into my commit of the notebook to your branch.
I made some minor changes to simplify your two masking steps. The cloud masking outputs are unchanged. I updated the land masking to make it dynamically mask each time step against itself which will hopefully account for changes in tide and inland flooding as relevant at other locations across Australia.
This PR also motivated an update to the dea_tools repo meaning your notebook now includes a direct calculation of NDTI and your subplots can now be labelled with their timesteps to support users to select the images for their animations.
I hope you're supportive of these changes. Feel free to recommend reverting to the original code, especially in the case of the masking. Otherwise, this notebook is a great addition to the resources in the repo and is ready to be added once the related dea_tools PR is approved. (I'll jump in and finalise the approval of this PR once the other is approved, unless I hear from you first!)
Thanks again!
|
That all sounds great! I think every change you have made has enhanced the code. I look forward to seeing it up on the sandbox. |
|
Yep, had a quick look too and this looks awesome - fantastic work @Joram95 and great changes @erialC-P! I have a few tiny suggestions (mainly around consistency with other notebooks, and a change to reduce the file size of the notebook) so I'll do a super quick review too once Claire's stuff is all integrated 😃 |
Proposed real world example notebook "Turbidity_animated_timeseries" created by Joram Downes
Proposed changes
This notebook allows the user to create a time series animation depicting a turbidity plume at the Murray Mouth and the surrounding coastline. The animations contrast Sentinel-2 imagery taken from early-2022 to mid-2023; this coincides with the flood of a generation, peaking in mid-January 2023, leading to a stark decline in water quality within the vicinity. This example demonstrates how to:
Closes issues (optional)
Checklist (replace
[ ]with[x]to check off)Load packagesGeneral advice)jupyterlab_code_formattertool can be used to format code cells to a consistent style: select each code cell, then clickEditand then one of theApply X Formatteroptions (YAPForBlackare recommended).NCIandDEA Sandbox(flag if not working as part of PR and ask for help to solve if needed)Notebook currently compatible with the NCI|DEA Sandbox environment onlyline below the notebook title to reflect the environments the notebook is compatible with