trias-project / risk-modelling-and-mapping Goto Github PK

@amyjsdavis will add Geotiffs of risk maps to GitHub for website testing

@SoVDH and @timadriaens : Here are a complete sets of PDFs for Symphyotrichum lanceolatum (GBIF taxon key:9202318) with a consistent set of file names. In my notes, I have written that you wanted to keep these simple. The scale is different for the difference maps as otherwise they would be hard to read. For the difference maps, a positive number represents increased risk, a negative number represents decreased risk.

9202318_rcp45_conf.pdf
9202318_rcp45_diff.pdfThe
9202318_rcp85.pdf
9202318_rcp85_conf.pdf
[Uploading 9202318_hist.pdf…]

9202318_rcp45.pdf

9202318_hist_conf.pdf
9202318_rcp_85_diff.pdf
9202318_rcp26.pdf
9202318_rcp26_conf.pdf
9202318_rcp26_diff.pdf

As @amyjsdavis noticed, it's time to move some functions she uses in her modelling to trias package.
This means:

1. make functions robust
2. document them
3. test them

First thing to do, @amyjsdavis, is list them before deciding which ones are worth to be moved and which ones are just very small help functions which can be left in the Rmd.

I can provide assistance on all these steps.

@amyjsdavis @DiederikStrubbe as discussed during the core group meeting, here are some ideas for visualisation of the usual model output (on top of projections for current climate and rcp scenarios 2.6 and 4.5) that could serve the risk assessment of alien species in Belgium, based on Beckmann et al. 2019 Projection of climatic suitability for Pycnonotus cafer establishment. Unpublished. (SDMs prepared for the EC PRA project, based on the method of Chapman et al. 2019.

• map with uncertainty on model predictions for Belgium (standard deviation in projected suitability). Example:

• bar plots showing the proportion of grid cells classified as suitable under current climate and projected climate. For Belgium, I would use Atlantic and continental bioregion, as well as the Belgian ecoregions. This can e.g. inform areas for targeted surveillance. Example:

but then with these regions:

• bar plots showing the proportion of grid cells classified as suitable under current climate and projected climate inside and outside Natura 2000. This could provide support to the risk assessor on questions relating to occurrence in protected areas. To do this, you can use the grid codes prepared by @damianooldoni who did an intersect with the protected areas (see #54)

For the EU PRA's, projections were classified into suitable and unsuitable regions using the ‘minROCdist’ method, which minimizes the distance between the ROC plot and the upper left corner of the plot (point (0,1)). No idea how that works though, perhaps there are other methods.

Further ideas:

• a limiting factor map following [Elith et al. (2010)](Elith J, Kearney M, Phillips S (2010) The art of modelling range-shifting species. Methods in Ecology and Evolution, 1, 330-342). Such map shows the most strongly limiting factor for establishment estimated by the model. I think this one is less 'important' for PRA-ing but it is informative.

• tabular summaries of variable importance, an example:

I don't think that describing the file naming of the output maps in a README pdf stored in the output directory is the best idea. I think this information suits perfectly the repo's README.
@peterdesmet, @amyjsdavis, what do you think about it?

Based on idea of @peterdesmet while discussing on this topic with @amyjsdavis and I.

All figures should be organized as follows:

data/processed/taxonKey/figure_name.extension

where

• taxonKey is the taxon key as mentioned in GBIF and data cube
• figure_name should include the type of model consistently

Any further suggestion, question or thoughts are welcome.

This is just a note. @timadriaens: could you please modify title and body of this issue based on what you jsut showed during the workflow meeting of today Oct 13? Please provide examples that @amyjsdavis can work on. I can assist her if needed.

Is there a reason of using sp as you are using sf as well.

Solve all small issues to make Rmd a nice html. This issue can be followed by many small issues/questions.
Again, I can provide some help on this process.

• @fjasteen can you link the paper and its appendixes here?
• @soriadelva goes through appendixes
• @soriadelva creates new scripts based on appendixes
• @soriadelva expands documentation
• @soriadelva @SanderDevisscher troubleshoot to get scripts working for test species (#18)
• ...

This is the file naming protocol I am proposing for the different types of maps for Belgium that will be output by the risk modelling and mapping code:

be_[GBIF_taxonKey][climate scenario][map type].tif (note that "be" indicates " Belgium" and the file format is always a GeoTiff)

Brackets indicate values to be assigned as follows:
[GBIF_taxonKey]== GBIF taxon id corresponding to the modelled species and used in data download and occurrence data cube.

[climate scenario]== can be 1 of 4 values as shown below:

[map type]== can be 1 of 3 values (including null) as shown below:

Example for Ailanthus altissima. GBIF taxon key is 3190653

be_3190653_hist.tif
be_3190653_rcp26.tif
be_3190653_rcp45.tif
be_3190653_rcp85.tif
be_3190653_rcp26_diff.tif
be_3190653_rcp45_diff.tif
be_3190653_rcp85_diff.tif
be_3190653_hist_conf.tif
be_3190653_rcp26_conf.tif
be_3190653_rcp45_conf.tif
be_3190653_rcp85_conf.tif

A question: where is the pra_mammals.csv file you are using in https://github.com/trias-project/risk-modelling-and-mapping/blob/master/src/global_download.Rmd#L43

It seems you are working on a local directory, ././modelling which is not synced to this repository. This should be avoided as it makes working with GitHub way less powerful than it could be. And the files in this repo become easily outdated.

Please, contact me or @peterdesmet if you have problems performing the "git jump" :jump
Notice that if you are not confident to share some files as they are not ready for being shared, you can list them in the .gitignore and they will not be pushed to remote (= this repo).

To improve repeatability the paths should be changed from absolute to relative.
Please, consider to use here package which is very useful as it allows finding paths from project root directory even if files are in src or if Rmd's are knitted.

Example of absolute path: https://github.com/trias-project/risk-modelling-and-mapping/blob/master/src/trias_sdm.Rmd#L47

To be rewritten as:

global <- read.csv(file = here("data", "external", "PRA", "PRA_mammals", "gbif_speciesFiles", gbif_filename))

I can assist if needed.

This issue comes from @amyjsdavis 's reply, #1 (comment).
Indeed, big files cannot be set in GitHub.

1. If data are already published on internet, a link to the source is sufficient.
2. If data are not yet published, they need to be published. You can publish them via Zenodo with label dataset and open access. In this way you can document, archive and, most of all, cite them! Example: https://zenodo.org/record/1288628.
3. If data cannot be published due to copyright restrictions, then a solution should be found during the Core Group meeting.

You can always ask @peterdesmet or me for advices.

Frederique can you provide 2 species that meet the following prerequisites ?:

1. The species has a limited number of worldwide occurrences on gbif
2. The species has some occurrences in europe on gbif
3. The species is relevant for your work

Shouldn't the content of PRA_processingSteps.txt be moved to the README.md?