ua-snap / alaska-wildfires Goto Github PK

If you use the form input field to enter a community name, select that community, then enter another community name, several errors show up in the console. The first error reads like this:

Error in callback for watcher "selected": "TypeError: Cannot read properties of null (reading 'latitude')"

Prevent this from happening:

STR -- select Baranof Warm Springs from community selector with console open. Then try to pick other places.

The idea here is that there's some layers which look/work well in combination or at different zoom levels, but it's not obvious to a casual user of the map how to activate/deactivate layers to get effective visualizations. We could explore making the map automatically show/hide layers (bonus points for transitional transparency?) to produce a really interesting map at any zoom level (0-6). A button to disable the auto show/hide should be available, too.

This needs to be fleshed out more, but roughly... when you're zoomed way in on the map, show the landscape classifications + hotspots + historical fire perimeters. Zoomed a bit farther out, lightning / current fire danger are interesting to see.

Pulling a clean copy of the data into GeoServer + correcting the symbology would be a good maintenance step; Kyle has made an improved version of these (see in Slack).

(iPhone/Safari)

• Title font way too big ("Alaska Wildfire Explorer")
• UAF logo is in the wrong place in footer

iOS/Safari,

When you scroll the page it's fine until you hit the map; then the scroll applies to the map, not the page. Fix this by detecting mobile + disabling mouseScroll in the right way for the map (make it require 2 fingers, usually.). Put a note right above the map (RIGHT above it when in landscape mode) to notify users.

Per Carolyn:

Minor text change suggestion from me—change:

The Wildfire Explorer shows a variety of data that helps users understand Alaska’s fire landscape. It is not designed for fire management decision-making by agencies or individuals.

TO

The Wildfire Explorer shows data that aids understanding of Alaska’s fire landscape. It is not designed for fire management decision-making.

Leaflet/Leaflet#8068

New feature which would work like this:

• User clicks on a point on the map
• App fires off request to SNAP data API for all data for that pixel
• A popup is shown containing the info for that specific point.

This doesn't have to be awesome, it's more to test the API integration.

Currently, if the user clicks a point outside of Alaska, it will:

• Zoom in on the map
• Not clear the air quality table, if a location within Alaska was clicked previously

Ideally, it'd be better to not zoom in, and clear the air quality table or replace it with a "No air quality data for this location" message. And maybe even replace the cursor with a "no entry" sign similar to our Fish & Fire app. But this all requires a lot more work, so probably something for future consideration.

From the AICC API, we can get a kind of "last status update" field for each fire. Grab this and show it as part of the info popup when you click on a fire.

It'd be nice if every legend for all data layers also indicated "where do I go to get these data?"

The data feed from the AICC rolls over to the new year, so there's no fires shown (accurate, but uninteresting). Indicate this on the site (hide map / everything and just have a "this site will be reactivated in April this year").

Right now we're using some layers from GeoServer; we can point instead at Rasdaman.

This would be in a future update, also see #7 for a possible x-reference/dupe.

• Change the name of “Projected relative flammability” to “Projected Flammability”
• Add another layer, listed between “historical fire perimeters” and “projected flammability”. This will be “historical modeled flammability” and will be the flammability data for the 1900s
• For “historical modeled flammability” the text could read “This layer shows output from ALFRESCO, a computer model that simulates the responses of Northern vegetation to climate change. Darker colors mean a greater chance of a region burning. These modeled data for the previous century allow for comparison between that century and this one, but do not necessarily match historical fire perimeters.”
• For “Projected flammability” perhaps slightly edit the text to say, “This layer shows output from ALFRESCO, a computer model that simulates the responses of Northern vegetation to climate change. Darker colors mean a greater chance of a region burning. These model projections for 2000–2099 can be useful for planning, particularly when compared to historical flammability and historical fires, but they can’t predict which specific places will burn.”

The publicPath is set properly in the vue.config.js file, but we're not necessarily using that properly everywhere for referencing image paths. It's a little tricky to sort out (confusing!) but that should be done so all images are correctly referencing their sources. The work to do here includes:

• Set the public path once only
• Use correct paths with all image references in HTML, JS and CSS (whatever that needs to be, possibly with webpack-specific prefixes)
• Ensure that images that are dependencies of components live in src/assets and that global/design images live in public/images. (probably. perhaps assets isn't the right place?)

Better to do this all at once instead of JsPrettier fixes sneaking in piecemeal through other commits.

We can compose some summary statistics and include that as part of the introduction as a dynamic blurb: "As of (date), there are (x) active fires which have burned a total of (y) acres. For the most current information, visit the AICC." < roughly.

i.e. "Land cover types, 2015" "Historical fire perimeters, 19xx-2022" etc. Line breaking will be a thing, need to iterate on that.

Be sure and scan the repo for 2021 lurking!

It's NCAR-CCSM4 / RCP 8.5. Add this to the layer legend.

For each legend, add some kind of outbound link to a data source!