woodwellrisk / viewer Goto Github PK

Home Page: https://woodwellrisk.github.io/viewer/

License: Other

JavaScript 100.00%

viewer's Introduction

Woodwell Risk data viewer

The purpose of this site is to increase public access to climate risk data while showcasing some of the datasets the Risk group uses to communicate climate risk to the communities and external partners it works with.

Data sources

Water stress

Drought index [NEX-GDDP-CMIP6]
Annual precipitation [NEX-GDDP-CMIP6]

Wildfires

Wildfire danger days [NEX-GDDP-CMIP6]

Temperature

Annual mean temperature [NEX-GDDP-CMIP6]
Hot days (days over 90°F) [NEX-GDDP-CMIP6]
Lethal heat [Woodwell]
Warm nights (nights over 20°C) [NEX-GDDP-CMIP6]

Coastal risk

Sea level rise depth value [NASA, Hammond et al. (2021)]
Tropical cyclone return period [STORM]

Processing steps

Vector data

Our boundary files came from Natural Earth. So far, we have used:

10m land and ocean boundaries [link]
10m country- and state-level boundaries [link].

All SHP files were converted to GeoJSON format in GeoPandas. From there, we used Tippecanoe to convert the GeoJSON files to Mapbox .mbtiles format and used the Mapbox tool mbutil to convert those tiles to .pbf format.

Raster data

For back-end data analysis/transformation of NetCDF and TIF files, we used Python and R. Those rasters were then converted to Zarr pyramids using CarbonPlan's ndpyramid package.

Acknowledgements

This site's interface and functionality rely heavily on code developed by CarbonPlan. Specifically, we used the maps, components, and layouts libraries. We took inspiration from CarbonPlan's forest risks code repository to create an updated and modified user interface for this data viewer. You can read more about CarbonPlan's research and software development work here.

viewer's People

Contributors

Stargazers

Watchers

viewer's Issues

Single value “region picker”

It would be nice to query an individual grid cell and get its value, rather than the default behavior of the region picker that takes the spatial average.

Slider input boxes

Add input box for exact min/max values in the colorbar sliders

Coastline is too detailed

Can we test how displaying a simpler coastline looks like? The "issue" is that, when zoomed out, the current coastline makes certain regions with complex coasts (e.g. Tierra del Fuego, Western Canada, Scandinavian Peninsula) look too "crowded", and hence, too white (black) under the dark (light) mode.

Compute data by vector layer

Look into computing averages / distributions of values for any vector layer, not just the region picker circle boundaries. This could be useful if we wanted to overlay a layer and be able to compute statistics as users hover over them.

Data viz charts

Add distribution plots, data values through time / warming level instead of plain averages when the region picker is opened

Custom colormaps

Either creating our own using CarbonPlan's colormaps package, or possibly importing them from Matplotlib.

Change chart values based on slider min/max

It would be nice to be able to dynamically change the min / max bounds of the x-axis on the bar chart with the slider that controls the colormap min / max when the <RegionPicker /> component is in use.

This isn't that hard to do, but one big problem is that the colormap slider doesn't actually remove values from the map. Instead, it hides them (for sequential colormaps) or changes their color (for diverging colormaps). This can be verified by using the <RegionPicker /> while changing the slider values.

So we either would need to be able to change / remove values from the raster itself or somehow rebin the existing raster values based on the min / max values from the slider.

Group data together into single zarrs

This will make the site render even faster, remove the flicker when switching between warming levels for a specific climate risk, and make it easier to those compare warming levels without resetting the colorbar limits. However, implementing this feature will mean rewriting significant parts of the code.