What is this about?

The Flying-Fox Heat Stress Forecaster has been substantially upgraded and is operational again!

The Forecaster now draws from weather forecast grids from the Australian Bureau of Meteorology’s Australian Digital Forecast Database, together with flying-fox camp and count data from the new Australian Flying-Fox Monitor (AFFM). 

​These upgrades allow the Forecaster to provide heat-stress forecasts for all known flying-fox camps in Australia up to seven days in advance, substantially extending the previous camp coverage and three-day forecast period.
Picture
Temperatures higher than 42°C (107.6°F) can kill flying-foxes ('fruit bats'; Pteropus sp.) at sometimes biblical scales. For example, on just one extremely hot day in January 2014, at least 45,500 flying-foxes died in 52 camps in southeast Queensland (Australia), wiping out about half of all black flying-foxes (Pteropus alecto) that were present in the affected area. More recently in November 2018, approximately 23,000 spectacled flying-foxes (P. conspicillatus) died during a single heat event in Far North Queensland. These deaths represented about one-third of Australia's entire spectacled flying-fox population, raising serious concerns regarding the conservation of this now-endangered species.  

To aid wildlife rehabilitators, land managers, and other stakeholders in coping with such events, our team from Western Sydney University and the University of Melbourne has developed a flying-fox heat stress forecaster. The forecaster predicts the camp (roost) locations where flying-foxes are likely to experience extreme heat, seven days into the future​.

​The Forecaster uses seven-day weather forecast grids from the Australian Bureau of Meteorology’s Australian Digital Forecast Database, together with flying-fox camp and count data from the new Australian Flying-Fox Monitor (AFFM). For details on the scientific basis and evaluation of the original Forecaster, see our paper by Ratnayake et al. (2019) in Animal Conservation. 

​The outputs of the forecaster include up-to-date maps of affected areas, hourly temperature profiles of affected camps, and a list that prioritises known camps by date, maximum temperature and the number of flying-foxes present. These are designed to direct stakeholders towards where and when flying-fox die-offs are most likely to occur, to facilitate efficient allocation of resources and help minimise impacts on human communities and flying-foxes alike.
 Heat stress alert status =
LOW =  heat stress unlikely
MEDIUM =  heat stress likely  (T ≥ 38°C)
HIGH =  mortality likely (T ≥ 42°C)
We need to learn more!
If you have any information about how flying-fox camps fared in recent heatwaves then please fill out our simple data form.

Latest temperature forecasts for Australia

Seven-day flying-fox heat-stress forecast animation
Animated GIF showing the current and predicted temperatures on the Australian continent within the three-day forecasting horizon. Predictions are based on computer model weather data from the Australian Bureau of Meteorology, and flying-fox camp locations (•​) are derived from the Australian Flying-Fox Monitor.

Areas on high alert (T ≥ 42°C)
48-hour flying-fox heat stress alert map
72-hour flying-fox heat stress alert map
96-hour flying-fox heat stress alert map
120-hour flying-fox heat stress alert map
144-hour flying-fox heat stress alert map
168-hour flying-fox heat stress alert map
Daily maps showing areas where temperatures are predicted to exceed 42°C in red. The maps can be loaded into Google Earth Pro (or similar) via their associated KMZ links.

Daily list of flying-fox camps where heat stress (T ≥ 38°C) or mortality (T ≥ 42°C) is likely
Note that:
  • Camps will be more at risk if they a) are subjected to higher temperatures, and b) have more flying-foxes present.  
  • To help prioritise camps that are likely affected by the highest temperatures first, camps are grouped by Date and ranked by Maximum Temperature. 
  • ​​To help prioritise camps that contain the most flying-foxes, the list also includes a Size.Class column, representing the number of individuals that is estimated to be using the camp at present (0=0; 1=1-499; 2=500-2,499; 3=2,500-9,999; 4=10,000-15,999; 5=16,000-49,999; 6≥50,000; blank cells means no count is available for that camp).
  • If no camps are forecasted to exceed 38°C, the file will read 'No heat stress is expected in camps over the current forecasting period' (in that case you can view an example output here).

Temperature profiles of camps where heat stress is likely to be worst

​Legend:   --- 38°C (heat stress threshold)       ---  42°C (mortality threshold)      o hours at ≥ 42°C (mortality  threshold)      --- twelve hour periods       --- one hour periods
Note that:
  • Here camps are ranked by maximum temperature. This is to help prioritise camps that are likely affected worst.
  • If no camps are forecasted to exceed 38°C, the file will read 'No heat stress is expected in camps over the current forecasting period' (in that case you can view an example output here).

Background information

 Flying-foxes and extreme heat events

Flying-foxes are charismatic, very large species of bat that forage by night and roost by day in arboreal roosts that can comprise tens of thousands of individuals. They provide important ecosystem services, including pollination of wild and cultivated crops and seed dispersal (e.g. Fujita & Tuttle 1991). However, they are exposed to threatening anthropogenic factors (Mickleburgh et al 1992), the most serious of which are  loss of foraging and roosting habitat (e.g. Tidemann 1999), direct killing of animals in orchards and harassment and destruction of roosts (e.g. Tidemann & Vardon 1997; Welbergen & Eby 2015). 

Our research has shown that temperatures exceeding 42ºC cause mass mortality among flying-foxes at landscape scales, with profound effects on the behaviour and ecology of the species. These extreme temperature events affect some flying-fox species more than others, with the more tropical P. alecto showing higher rates of mortality than the more temperate P. poliocephalus. The events are also age-class and sex specific, with dependent young and lactating females predominantly affected. 

So far we have found evidence of at least 39 flying-fox die-off events in Australia, with the earliest dating back to 1791, but the majority (35) occurring since 1994. At least 28 events had a death toll in excess of 1,000. In these 28 substantial events, over 233,300 flying-foxes died, with >200,000 deaths recorded in the last decade alone.​Currently, flying-fox die-offs occur between 0-5 times per year, but these events are expected to escalate under climate change, with disturbing implications for flying-foxes and other organisms in Australia and beyond.​
Flying-foxes show a predictable sequence of thermoregulatory behaviours with rising temperatures (clockwise: wing fanning, clustering, salivating, panting). Beyond 42ºC, thermoregulatory mechanisms become overwhelmed and this is soon followed by death (Welbergen et al 2008).
The flying-fox heat stress forecaster
The Flying-Fox Heat Stress Forecaster uses hourly air-temperature forecasts from the Bureau of Meteorology’s Australian Digital Forecast Database to provide camp-level forecasts up to seven days ahead.

Flying-fox camp locations, species and count records are obtained from the Australian Flying-Fox Monitor (AFFM). The Forecaster displays the latest available count and the date on which it was made, allowing users to interpret forecast heat-stress risk in the context of both the number of animals last recorded and the currency of that information. By connecting camp records, short-term heat-risk forecasts and field-based response information within a common spatial framework, the wider AFFM architecture also supports more coordinated monitoring and response.

The weather and AFFM data sources are checked independently during each hourly processing cycle. When a new Bureau of Meteorology forecast or updated camp information becomes available, the forecast grids are processed in R, linked to the current camp frame and validated before replacing the previously accepted forecast. The Forecaster identifies camps predicted to experience air temperatures at or above 38°C, when heat stress becomes likely, and at or above 42°C, when mortality becomes likely.
​Animated GIF showing forecasted temperatures in eastern Australia, ​mapped from Australian Bureau of Meteorology computer model data. 
​The scientific basis and original evaluation of the Forecaster are described by Ratnayake et al. (2019) in Animal Conservation. The operational system has since been substantially revised and updated, including its weather data source, forecast horizon—which has increased from three to seven days—camp-data source, and automated refresh process and outputs.

​Useful links

Extreme heat events vs. flying-foxes:
Management responses:
​Health and safety first!
NSW Government (Heat-related illness including heatstroke); NSW Government (Beat the heat – information for health professionals); NSW Government (Beat the heat – How to stay healthy in the heat); Life Saving Victoria (Heatstroke); Northern Territory Government (Guidelines to prevent heat stress); Smartplay & SunSmart (UV exposure and heat illness guide); RSPCA (What can I do in hot weather to prevent heatstroke in my pet?); SA Government (Extreme heat)
For general maps of impending heat waves with implications for human health and infrastructure, see the Heatwave Service from the Bureau of Meteorology.  ​​

Attributions & acknowledgements

Conception: Justin Welbergen, Luke Shoo, David Karoly, Mike Kearney
Scripts: Justin Welbergen, Mike Kearney, Himali Ratnayake, Pallavi Govekar, Stewart MacDonald
​Web design: Justin Welbergen, Himali Ratnayake

Many thanks to Kamal Puri, Robin Bowen and Tan Le from the Australian Bureau of Meteorology, and Adam McKeown and David Westcott from CSIRO for their help with the earlier version of the forecaster (2016-2025).​
​

Necessary disclaimer

Our flying-fox heat stress forecasts are based on best-practice landscape-scale weather predictions from computer model data. Nevertheless, the predictions may under- or overestimate maximum temperatures, especially at the small scale of individual flying-fox camps. In addition, they may exclude camps that are currently occupied, and, conversely, include camps that are currently unoccupied. ​We cannot accept any responsibility for the loss or damage that may result from reliance on, or the use of, any information contained on this webpage.
​THIS REMAINS A WORK IN PROGRESS!
Please CONTACT US with any suggestions for improvements
​[for known issues see here]​