Bird observation records retrieved from the Atlas of Living Australia's (ALA) database have been filtered to exclude records with obvious issues. BirdLife Australia provided detailed species range information that allows us to classify records as core habitat, introduced, historic, vagrant, irruptive, or doubtful. Observation records that fall outside these ranges are marked as unclassified, and are assumed to be valid for modelling. However, we need input from birdwatchers and experts to classify these records so as to improve the models. Additionally, there may be pockets of unsuitable climate space within core habitat, which also requires vetting.

The initial Birdlife classification for observations is compared against classifications collected from birdwatchers and other knowledgeable visitors to the Edgar site. Edgar tracks contentious records, but generally a vetting entered by a logged-in site user will be considered accurate and changes the derived classification of an observation.

Doubtful records and records that are considered historic, irruptive, or vagrant are not used to model climate suitability for a species. Only species with >20 unique location records are modeled.

We used climate data from the Australian Water Availability Project (AWAP) to caculate important climate variables such as current annual mean temperature, temperature seasonality, and annual precipitation. We calculated likely future climate for each climate variable considered important to bird distributions using all 18 global climate models (GCMs) for all RCP scenarios (RCP2.6, RCP4.5, RCP6, RCP8.5) at 8 time steps between 2015 and 2085.

Climate suitabiliy maps for a species represent what scientists call a Species distribution model. These model shows the relationship between where species have been observed and the climate at that location. Edgar shows all potentially suitable climate space, even if the species has not been observed there.

Edgar continues to collect new observations from ALA and new vettings from site visitors, and as species observations are changed and refined, Edgar regenerates the species' climate maps using the new data. The most up-to-date maps of climate suitability are displayed on Edgar.

Unclassified

Observation records that have not been classified within any other category. With nothing else to go on, these records are assumed to be valid and are included in the modelling. Please improve future models by classifying these records!

Core

An observation record that falls within any area essential to survival of the species. This includes breeding grounds, stop-off points for migratory birds, and locations that are visited seasonally.

Historic

The species has not been observed at this location since 1975. We use the cutoff date of 1975 because climate change is underway, and this is reflected in species distribution changes. If a bird species has not been observed at a location since 1975, we cannot assume that the climate at that location has remained suitable for the species.

Irruptive

Observations at a location represent a dramatic, irregular migration of many birds away from their core range. These irruptions may occur in cycles over many years, or they may be much more unpredictable.

Vagrant

Individuals of a species have appeared well outside their core range. Escapees that have not established a population are included within this definition on this site.

Introduced

A native or non-native species has established a population within an area well outside their native core range.

Doubtful

It is very unlikely that this species was observed at this location.

Representative Concentration Pathways are four green house gas trajectories adopted by the IPCC for it's fifth Assessment Report. The RCPs are not new, fully integrated scenarios (i.e., they are not a complete package of socioeconomic, emissions, and climate projections). Here, we have used climate layers generated by Jeremy VanDerWal that have been derived from the following RCPs:

RCP2.6

The RCP2.6 emission pathway represents scenarios leading to very low greenhouse gas concentration levels. A peak in in greenhouse gas levels is reached mid-century after which levels subsequently decline. Emissions reduce substantially over time.

Mean temperature increase for Australia: 0.91 degrees by 2085

RCP4.5

The RCP4.5 emission pathway is a stabilization scenario where emissions stabilise before 2100 by employment of a range of technologies and strategies for reducing greenhouse gas emissions.

Mean temperature increase for Australia: 1.83 degrees by 2085

RCP6

The RCP6 emission pathway is a stabilization scenario where emissions stabilise after 2100 by employment of a range of technologies and strategies for reducing greenhouse gas emissions.

Mean temperature increase for Australia: 2.29 degrees by 2085

RCP8.5

The RCP8.5 is characterized by increasing greenhouse gas emissions over time leading to high greenhouse gas concentration levels.

Mean temperature increase for Australia: 3.78 degrees by 2085

Edgar performs three roles:

1. show observation records and current climate suitability for a species
2. show future climate suitability
3. vet observation records.

Edgar starts by showing you a search box across the top of a map. If you are alreadying viewing a species, you can switch back to the search box species by clicking the 'change species' button near the species name.

Type a species of your choice into the search box. As you type Edgar will pop up a list suggesting matching species; when you see the species you are looking for, click its name (or use your arrow and Enter keys to select it).

While viewing a species, bird observations are clustered so that a single coloured "dot" summaries the observations across an area. A cluster dot is drawn larger to represent a high number of observations in that area, and smaller to represent fewer observations.

As you zoom into an area, cluster dots are re-drawn so that each dot covers a smaller area. When you zoom in far enough, observations un-cluster to represent the recorded location of bird sightings.

Select the species as described above, then click the "see future projections" button near the top right corner of your map. Edgar will show you a new set of tools down the right side of your screen, and start loading map projections. This will take a few seconds, or longer on slower internet connections.

Once the future climate suitability maps have loaded, click the green-circled play button to watch how the suitability will shift into the future. You can also drag the slider to examine any modelled year from 2015 to 2085 in ten year steps.

You can view future climate suitability under an alternative emission scenario by selecting the scenario in the right column.

To vet observation records, you will need a user account. To help reduce the number of logins you might need to create and remember, Edgar shares the user account system from the Atlas of Living Australia.

Click the 'Log In' button at the top right corner of the screen. This will take you to the Atlas of Living Australia (ALA) website. If you already have a username and password, log in; if you do not, click the registration link to register. Registration with ALA only requires your name or alias, and a valid email address.

Once logged in to ALA, you will be returned to the Edgar site. Enter the species you would like to vet into the search box, then click the 'vet this species' at the top of the right tool bar. Edgar will show you a new set of tools down the right side of your screen, including a 'vet observations' tool below the classification legend.

Recording a vetting is a three step process:

Step 1: select observations

To record a vetting, click 'Select Observations' and either:

• click an observation or cluster of observations, or
• click-and-drag a selection box over multiple observations or clusters.

Selected areas will be highlighted with orange boxes. To de-select these areas, click 'Clear Selection', then choose 'Select Observations' to start again.

Note: You can select multiple areas across Australia and give all selected areas a single classification. For example, you might wish to classify multiple records offshore or in the central desert as 'doubtful'.

Step 2: enter your opinion

Having selected the observations you want to re-classify, choose a new classification from the drop-down menu in the 'vet observations' toolbox. You may add details about your reasons for your classification choice if you like.

Step 3: save

Finally, click 'Save this Vetting' to apply your classification. A confirmation popup will appear and the selected area will be highlighted according to the classification you provided.

When you have finished vetting one species, you can begin vetting another species by clicking 'Change species'.

Why can't I find the species I'm interested in?

We filter the species list to only include species for which we have observations. It's possible that the Atlas of Living Australia has no observation records for the species you're looking for.

Why are there no records for this species?

We display all bird observation records retrieved from ALA so that we can classify those records and send that classification back to ALA. If no records are shown for a species, it means that all records for that species have been classified as 'doubtful'. If you wish to see the doubtful records, register an account with ALA and log in to Edgar. You may then see and vet the records.

Why is there no climate suitability map for some species? (ie. Why has a species not been modelled?)

We can only model species that have 20 or more observations at unique locations at a 5km resolution.

Why does the climate suitability map show suitable areas outside of where a species has been observed? (ie. Species observed only in Victoria has suitable climate space in Western Australia)

Edgar shows all potentially suitable climate space, even if the species has not been observed there. This is because managers may need to take all potentially suitable climate into consideration when developing conservation plans for a species into the future.

How do I obtain accurate observation records for a species?

Most data available for download on Edgar is accurate, however some location observations for some species are considered sensitive information and we cannot release it either visually (displayed on the map), or as downloadable data. If you require accurate observation records for a species, contact ALA.

How do I report a bug/suggest a new feature for Edgar?

Email Jeremy VanDerWal.

Can I use Edgar / Edgar's data / pictures from Edgar / text from Edgar in my project?

Edgar's source code is released under a BSD-style licence and available from github.

Edgar's content, including Edgar's documentation, is licenced using CC-BY, which essentially means you can do whatever you like with Edgar's data, as long as you credit it properly.

via email:

jeremy.vanderwal@jcu.edu.au

by post:

Centre for Tropical Biodiversity & Climate Change Research
School of Marine and Tropical Biology
James Cook University
Townsville, QLD 4811
Australia

Currently there is a general lack of engagement and knowledge transfer between professional researchers and end-users of research (general public, conservation managers, decision-makers, etc.). This is reflected in a general lack of acceptance and acknowledgement by the general public of the potential impacts of climate change. Indeed, the ABC reported 27 June 2011 that

"The Lowy Institute's annual poll asked about 1,000 people for their opinions ... The poll shows that there has been a steep fall in the number of Australians who think climate change is a serious problem which needs addressing now."

Research individuals or groups spend considerable time and effort to bring together species occurrence data, but substantial effort is still required and limitations exist with respect to a) the accuracy of the localities, removing only blatantly incorrect records outside known locality, or b) using painstaking manual processes whereby occurrence records are presented as hardcopy maps to "species experts." These experts then write metadata (e.g. provenance information), corrections, and other information about further records on the maps and return the comments for interpretation. This is a cumbersome, labour-intensive, and error-prone process which needs to be repeated for each project.

There is currently a scarcity of transparent online tools which integrate species distribution data, locality data with climate change scenarios in an integrated fashion which will facilitate the modelling of current and future species distributions based on climate scenarios.

The Edgar site provides a tool that reuses data available with the Atlas of Living Australia and the Tropical Data Hub to allow a broad range of end-users to:

• explore with the potential impacts of climate change on a wide range of species in Australia
• engage in improving our understanding of the species and the modelling of species distributions