From June 2019 to February 2020, Australia has faced fires with unprecedented timing and intensity. As of January 14, 2020, at least 27 million acres of land have been burned, causing an estimated 3–3.5 billion dollars in damages. As climate change has gained greater recognition in the media, many are wondering how much changing weather patterns are to blame.
In response to this speculation, however, there have been high rates of falsified information on the relationship between climate change and Australia’s fires circulating on the web. Fortunately, it’s Planet OS’s mission to make reputable weather data easily available. With hundreds of publicly accessible climate and environmental datasets on the Planet OS Datahub, our team saw this as a great opportunity to use these datasets to provide reliable, data-driven visualizations on the relationship between climate and Australia’s fires.
Australia’s Climate and Fire Risk
Using ECMWF ERA5, Bureau Of Meteorology (BOM) Daily Rain, Temperature And Vapour Pressure For Australia, and ECMWF ERA5 Fire Danger reanalysis datasets, Planet OS Data Integration Engineer, Eneli Toodu was able to create the below visualizations of fire weather index, temperature, and precipitation.
Fire Weather Index: 2019 and over the Past Decade
Using the ECMWF ERA5 Fire Danger reanalysis, Eneli was able to create the following visualizations of the Fire Weather Index in Australia. The Fire Weather Index (FWI)is a numeric rating of fire intensity, dependent on weather conditions. It functions as a strong indicator of fire danger because it contains both a component of fuel availability (drought conditions) and a measure of ease of spread. For more information on FWI, click here.
To the left, we see FWI illustrated over Australia on November 12th, 2019. This shows FWI conditions when the state of emergency was declared in the greater Sydney area and fires were rapidly spreading. Conditions are marked from “low” fire danger (green) to “extreme” (red). It should also be noted that FWI shows fire risk from weather conditions, not environmental conditions. Thus, regions that have weather conditions that create a high fire risk, but do not have burnable vegetation and/or built conditions will still register as “extreme” on the map.
Located above, Eneli was then able to create a yearly FWI for Australia in November from the years 2010 to 2019. From 2010 to 2019, there is what appears to be an increase in “extreme” FWI ratings from one year to the next. From this, we can see there were a greater amount of regions in Australia with weather conditions that were extremely conducive to fires over the nine years.
Next, we took a look at temperature averaged in all of Australia, and Victoria and New South Wales specifically. Victoria and New South Wales have been the worst affected by the fires, destroying more than 2,500 houses and forcing thousands to seek shelter elsewhere.
Below is a graph of temperature anomalies for Australia (yellow), Victoria (orange), and New South Wales (teal). Eneli generated this graph using the ECMWF ERA5 dataset. The Annual Temperature Anomalies are calculated from the mean temperature in each location from 1979 to 2019.
From this anomaly graph, we can see a trend in positive temperature anomalies (temperature anomalies with 0.1+°C above the mean) from 2013 to 2019. To take a closer look at this trend, Eneli then graphed the Yearly Mean Temperature for the three areas.
Both the anomaly and yearly average graphs show a linear positive trend towards warmer weather conditions for the past five years. While more data analytics and research are needed to determine clear, concrete relationships, the visualizations show notable patterns with the raw data.
Using the Bureau Of Meteorology (BOM) Daily Rain, Temperature And Vapour Pressure For Australia we then looked at precipitation in Australia.
As seen above, the BOM dataset has the capability to plot precipitation anomalies from 1900 to 2019; an over 119-year span of data. Since the extent of information visible above is so vast, a variety of notable trends becomes apparent. For example, with a quick glance for the years of 2016 onwards leading up to the 2019 fires, the precipitation anomalies were in the negatives. This means that millimeters of rainfall for those years were much lower than the average.
Although droughts in Australia are a regular event, one that we have analyzed since 2018, we see the severity of these climate patterns has increased through the more dramatic anomaly values. This becomes especially evident in the above anomaly graphs where the upper and lower peaks become significantly larger post-1950.
We can also see on these graphs that New South Wales and all of Australia had the hottest, driest year on record in 2019. These conditions undoubtedly left the continent especially vulnerable to severe fires.
The questions and answers one can have with data are infinite. In the case of Australia and the massive fires the continent experienced for the past seven months, climate and environmental datasets can pave the way to groundbreaking discoveries and solutions. There is an undeniable relationship between climate and the veracity and spread of bushfires, one that we hope will be addressed with proactivity and data-backed decision making.
Our team is eager to see how weather data is applied to better understand why the 2019 fires in Australia were so severe and what can be done from now on to protect communities and ecologies. Technology may hold the key, and perhaps with the integration of more accessible information, we can reduce destruction and keep communities safe.
The Planet OS Datahub makes it easy to build data-driven applications and analyses by providing consistent, programmatic access to high-quality datasets from the world’s leading providers. If you are interested in the types of resources that the Datahub provides, but cannot find the dataset you need, please don’t hesitate to contact us via Slack or email. Many of the datasets made available through the Planet OS Datahub have been at the request of our users. For more information check out the Planet OS Datahub.