Our new migrant bar charts can be accessed HERE. They will be released in batches, with the first few posted today (25 Aug).
Have you ever asked when you should expect to start seeing warblers and flycatchers make their way to Singapore, or when to start watching the skies for the fall raptor migration?
Our bar charts for rarities were released early this year, and we trust that it has been a useful tool for many in deciding when and where to look for mega targets. After making them public, we decided to dream bigger: why not do it for all migratory species in Singapore so that everybody can learn more? In the coming weeks, we will be launching bar charts for all migrant and vagrant species (not just rarities), illustrating the estimated number of individuals that are typically present in Singapore during different times of the migratory season. These charts will also highlight the peak weeks, as well as early and late dates for relevant species (only possible for species which do not have oversummering records).
Our charts will be rolled out and launched on our website progressively based on the typical arrival times of each species.
Data used is based on the period Jul 2012-Jun 2022, and will be updated at the end of every season so it remains relevant over time.
Use this page to access our bar charts, or keep reading below for a more detailed overview of how they are produced.
How we generate our estimates
For species in our database, this is relatively straightforward. We split each year into 52 weeks, starting from Jan 1, and count up the number of birds in each week according to records in our database.
It is a more involved process when it comes to estimating numbers for species not included in our database.
Briefly, the process includes:
- Grouping all the sightings into their corresponding weeks
- Using distance between sightings to determine whether two sightings may be linked (i.e. the same birds involved)
- Using time between the sightings grouped in step 2, to further determine whether they are linked
- Concluding by lumping those sightings which are linked, and splitting those which are not, to give an estimate of the number of birds present at any one point in time
This procedure is necessary to account for duplicated records as not every observer uses the same eBird “hotspots”, and some users may use “personal locations” to record sightings rather than hotspots. This may result in a distance of a few kilometres between eBird observations that might refer to the same bird.
At the same time, different species may require different methods of determining whether two sightings are of the same bird. A raptor which is passing through Singapore is unlikely to be seen on multiple days in a row – two Common Kestrels at Henderson Waves on consecutive days in October, for example, are likely different individuals. Our approach, which combines spatial and temporal methods to estimate the number of individuals of a species during each week, is relatively tolerant to these complicating factors.