Name of Project:
Proposal in one sentence:
Building a web app to allow users to quickly and easily visualize dClimate weather, climate, and forecast data, as well as analyses of this data.
Description of the project and what problem is it solving:
Visualization of weather, climate, and forecast data is an integral component to understanding this data. Research scientists, climatologists, forecasters, modelers, and data scientists - as well as anyone who derives products from this data - often want to quickly understand certain features about their data:
-What are long term trends?
-What are the extremes of the data?
-What is the value of current data?
-How far back does this data stretch?
-What is the coverage of this data? (spatial and temporal)
In order to answer these questions, individuals often have to do their own inhouse analysis (call the api, query the relevant data, write a python script, plot the data). This project intends to cut out this unnecessary work and make visualization and analysis easy and fast, allowing users to understand the data quickly. Furthermore, this will make this data accessible to non-technical audiences.
This project will be written exclusively in Python (using Dash and Plotly) and hosted on Heroku.
An MVP of this project is already available at dclimateviztest.herokuapp.com. The objectives for improvement of this MVP can be found below in grant deliverables.
Grant Deliverable 1 - Add functionality to plot second element
Grant Deliverable 2 - Add functionality to plot analyses (e.g. histogram, SMA, etc.)
Grant Deliverable 3 - Add other datasets (forecast, GHCN, etc.)
Grant Deliverable 4 - Improve UI/UX (time permitting)
Funding Requested: 1000 USD
Adam Filipovich (University of Toronto, Environment and Climate Change Canada)
Value add for Algovera community:
-gives quick access to dClimate data for Algovera community members
-facilitates understanding of dClimate data for algorithm creation
-community-created indicators, which are derived from API data and can be represented as time-series themselves, can be implemented into dClimateViz, fostering an ecosystem of transparency around climate data indicators