The central part of the infographic is the map of Russian regions. Colour saturation of each region denotes the number of registered wildfires per unit of regional area in a selected month. The month can be changed with the slider above the map. If you select the last month, you can see wildfires burning at the moment.
In addition, you can explore wildfire statistics since 2010. Below the map, the histogram shows the area covered by fire for each day. Fire distribution and burning area are displayed when you mouse over the histogram.
Wildfires are a powerful natural and anthropogenic factor significantly altering the condition of forests all over the world. Fires cause damage to the environment, economics and endanger people's lifes. For countries where forests cover a large area wildfires have become a big problem.
Annual reports highlight that the number of wildfires across the territory of Russian forests varies from 10,000 to 35,000. They are covering the area of 2.5 to 7.5 billion acres (1–3 million hectares) of densely populated regions; from 5 to 14 billion acres (2.0–5.5 million hectares) of unprotected and occasionally protected areas of Northern Siberia and Far East.
This project aims to provide live data regarding wildfires in Russia and observe the dynamics of firefighting in separate regions on the time scale of a few years. The main source of the data is NASA project EOS. You can read more about data sources in project details mentioned below.
This project requires data from four different external sources. All the data is combined, preprocessed and stored in Postgres database.
Live wildfire data source is NASA project EOS. It provides near real-time active fire data for users interested in monitoring and analysing a wide variety of natural and man-made phenomena.
Precise correspondence between fire spots and regions was created with Yandex.Geocoder.
Area and topographic data were obtained from Wikipedia.
I needed to overcome some difficulties during the development of the project.
The first one was the amount of data presented on NASA site. Since 2010, there have been registered two million wildfire disasters. I had to establish correspondence between each fire event and Russian region. Yandex service allows only 25 thousand requests per day. So I implemented approximate estimation of regions based on precisely classified points. If the point in question is surrounded with ten reliable ones those belong to the same region, it's assigned to that region. This approach was found as absolutely appropriate for the purpose of the project.
Client optimization was another pitfall I faced. I had to present all data without overloading the browser. Thus, my decision was to load statistics on demand. When user requests the data about specific day, it's loaded asynchronously and displayed as soon as possible. As a result, the application starts quickly and loads the data while user is reading a project summary and instructions.
You can get detailed wildfire information for any day from January 1st, 2010.
Make GET request to the following url:
For example, get the data for April 5th, 2015.
You can get details about any wildfire with known ID by making GET request to the following url:
Get the data for wildfire #673634.
To get the data in pure json you should make response with HTTP-header
Accept == 'application/json'.
You can get automatic wildfire alerts via Twitter.Твиты от @WildfireStats
You can download detailed wildfire information for any day in
Follow the link of such a scheme:
Try to download the data for March 15th, 2015.
Web interface is managed by Django backend. User interaction and animation is based on d3 library. All data miners are written in Python using requests and urllib2 which try to load fresh data every three hours.