trias-project / risk-maps Goto Github PK

Load data from GBIF's checklist? Or from Amy's files? To be investigated.

Request from May 6th meeting:

• Show more species, not only the ones we have GeoTiff for.
• Occurrence data would still originates in the same data cube (=show all occurrences from the cube)
• Occurrences would be filtered: only show 1976-2020
• Need a user-friendly way to announce to users which species has risk maps or not

0-1, description given by Amy at trias-project/risk-modelling-and-mapping#7

It is unclear what those colours represent. A legend should be added, with a unit and values.

Issue created to collect the overlays we want to be able to show along the risk maps.

Note: If possible, all shown data should be INSPIRE-compliant.

Themes:

• N2000
• Cities
• Rivers
• Ecoregions
• Belgian regions

Related technical question: how to show them in client only webapp: convet to GeoJSON? Possibility to show shapefiles? Should we reproject those prior?

@niconoe : I just pushed more maps to the GeoTiff folder. :-)

I'm not a big fan of the background layer. What is available? It is mainly cities and rivers that are interesting. Not really roads in my opinion.

I think would be nice if the application started with a loaded example, e.g. the first value from each dropdown (excluding "Ecoregions"), rather than:

The source datasets should be listed somewhere. I assume this will be the same ones as for the cube.
Ultimately a dynamic sources citation would be best, but a list will do for now.

@silenius: here is in a nutshell the envisioned development. Since the idea is that this work is at a later step embedded in a larger development you'll made for TriAS, I wanted to make sure the approach I take suits you. The plan at this stage is:

• A simple (if possible frontend-only) app that display the content of GeoTiff files over a Leaflet map of Belgium. There will be some usual overlays (ecoregions, N2000 zones, ...)
• Implemented using Vue.JS/TypeScript (Vue-CLI for a single page application - the written vue components can obviously be reused in other contexts-
• The app and UI would be pretty basic: a select to choose a given species, another select to choose a "map type" (there are multiple geotiffs generated per-species, for example to show the species progression according to various climate scenarii). From those two selects, we can infer a filename for the GeoTiff, and bam => on map.
• I'll use some frontend framework (probably Bootstrap, except if you have other suggestions/preferences) to have a basic page structure and nice form. It'll be limited, so it should be relatively easy to remove/replace them if needed when you do further integration.

Don't hesitate to comment!

Take inspiration from https://leafletjs.com/examples/choropleth/ ?

When selecting an area, all areas of that type should be highlighted, so it is clear that these are of the same type. E.g. three areas for polder, while currently only one is highlighted:

Hi, think it would be useful to give users the option to plot the actual distribution (from the cube of from gbif directly), so they can explore to what extent potential area in Belgium has already been invaded. Need to think about proper date cut-offs (e.g. from 2000), type of data to use (human observations or other).

That's always useful for traceability reasons (or just to check if a deployment was successful)

The files are formatted as ${this.publicPath}geotiffs/be_${this.speciesId}_${this.climateScenarioId}_${this.mapTypeId}.4326.tif
But for the scenario with rcp85_diff the filename has an extra underscore. So these tif files are not picked up by the tool.
I see the inconsistent naming already originates in the /source_data/geotiffs folder, so probably a bug in the script that creates these files?

For reference: We aim to integrate these risk maps in the alien-species portal

Let's write a script that parse a directory containing Amy's generated files and:

• copy them (converted to EPSG4326, to feed the webapp)
• write some JSON configuration file for the selection form (species name, GBIF IDs, ...)

(but works fine if we go to "none" in between)

@niconoe: @SoVDH noticed that maps for 5281901 (Campylopus introflexus) are not showing on the viewer. The geoTIFFS for this species are in the google drive folder you created for GeoTIFFs, but they are in a subfolder I created for my last upload called PRA_Plants_noUnverified. (This upload addresses the issue of unverified occurrences that @damianooldoni discovered last summer). This folder contains the final maps that should be used in the viewer, including the one for Campylopus.

I realize now that I created confusion by adding subfolders, and if you like, I will clean up the google drive folder leaving only the final version of the maps only while eliminating duplicates and subfolders. I wanted to check with you first. Thanks!

I don't know where you got them from, but the names of biogeographic districts (ecoregions) are not complete in Wallonia and do not refer to what is commonly used as far as I know.

• Condroz
• Fagne-Famenne-Calestienne
• Thiérache
• Ardenne
• Gaume et Lorraine

Also, Maybe worth having both FR and NL names for the users?

@peterdesmet : at a later stage, will you be available to make the web app looks better?

It will stay pretty small, and the toolbox is the same than Crow.

This hides the controls on smaller screens:

I would release it with a disclaimer. @amyjsdavis @SoVDH can you provide a short sentence that we can display at the top of the application. Something along the lines below:

The maps and models shown below are in an experimental state and should not be used for policy ...

Originally posted by @peterdesmet in #10 (comment)

It took me a while to notice that the name of the hovered overlay was shown on the right. It also disappears right away if you hover away. I think this could be improved by:

1. Showing all elements of an overlay on the side
2. Highlighting the element when it is hovered on the map and vice versa
3. Highlighting the element when it is clicked on the map and vice versa
4. Clicking an element (map or side) keeps it selected. This selection is undone by a) clicking it again, b) highlighting another area, c) clicking anywhere else on the map (including outside the full overlay)

Would it be possible to add as overlay the simple 'atlantic' vs 'continental domain' ? These are basic unit used for risk analysis and users will need them anyway.

It's currently probably still a test, but the styling of the overlays can be improved:

• Red highlight: good that it is transparent, would use other colour, e.g. TrIAS orange: #DC571D
• Borders: use full lines, not dotted
• Borders: use stronger colour than white (it is a bit hard to see, especially when zoomed out), e.g. TrIAS orange: #DC571D

The cube contains more, but the modelled data is using that time range.

(will plan a meeting soon for this)

• Allow species selection (GBIF taxon key in filename?)
• Allow to select a risk map type (multiple files per taxon)
• Improve visuals: color scale, allow to change transparency, ...
• Create a basic layout (Bootstrap-based?)
• ...

Hi All: I am sorry I missed this last month. I had asked Nico (while I was unaware of this thread) to only show only the occurrence data I used to make the risk models, because there is sometimes a big difference between the date I used and the occurrences that are shown in the risk mapping application. Now that I've read through this thread, it seems like the biggest differences in the occurrence data he is showing and the one I used for the risk modelling is time and excluding data based on coordinate uncertainty. In order to align the models with the historic climate data, only data from 1976 to 2005 are used in my models. I had prepared shapefiles that show the occurrence data used in the models for Belgium and that is ideally what I like to use since he can't filter based on coordinate uncertainty. The 2nd best option is to filter the data based on time. I think that would make the occurrence data closely approximate what was used in the risk models. What do you all think? @niconoe

Originally posted by @amyjsdavis in #21 (comment)

for the sake of having some more information on riparian and aquatic IAS, I think it would be nice to have some overlay maps regarding the hydrographical network. Rivers and canals people can see if they zoom in, but being able to put the river basins in Belgium on the risk map would be very informative. Ijzer, Schelde, Maas and Rein are of course very big delineations, it would be nice to have something a bit more detailed e.g. at the level of river basin management planning.

the least we can do is perhaps to provide the river basin units used by the Water Framework Directive but perhaps the subunits are more informative since they are more the level used for river management planning (and IAS management should be at that level).