This pioneering weather map from the Sydney Morning Herald 5Feb1877 has a list of 33 towns and locations in New South Wales where daily weather observations, rainfall, max & min temperature, barometric pressure etc were being recorded. 34 if you count Sydney Observatory Hill.
For many years I have been puzzled at the deterioration in BoM rainfall data more recent than about 1995 and this list of heritage NSW weather stations is a useful source to test my 2008 claim on a group of significant BoM stations choosen by somebody else. For each location name I have searched CDO for monthly
rainfall data from ~1870’s and for sites that closed decades ago to fill gaps. I have added a site or sites till I got data ~current to 2019. I have not cut and pasted every location into a spreadsheet to test for every missing year – that can wait until I have more time.
My spreadsheet records 64 stations used to build a full record for each of the 34 locations – many could be improved by adding data from the closest Airport AWS which might be further away and as I said in the above sentence there may be a few where missing years I have not eyeballed need hunting down from neighbours.
Here is my spreadsheet where marked in red are 22 examples of “post ~1995 gappy data”.
You can check the data for yourself near the bottom of the CDO page where 3: Get the Data invites you to
type the station number and hit the Get Data button. Easy. See examples of the data deterioration –
increasing gaps – for yourself. Modern AWS sites are not immune to data gaps.
With all the gaps in the rainfall data it makes me wonder the amount of calculation required to produce the BOM rainfall deficiency charts like these :
www.bom.gov.au/climate/drought/archive/20191008.drought2.lr.col.gif
Maybe locations that are included in the driest 18 months on record zones have very little continuous 18 month periods to compare to.
Terrific effort, wazz. This is the sort of research the BoM should be doing itself, and more importantly, it should then be fixing the problems identified. They can’t have it both ways – treating the historical data as a treasure trove for their bogus “homogenised” global warming series, but then letting those data deteriorate so badly so that we cannot even see the current numbers from the same sites.
Looking at all the gaps and deteriorations, I wonder if the automatic stations are really any improvement at all. In the old days it was actually a job to get and record the data, so somebody did it, and constantly made sure the equipment was working. Now that it has become an automated process, the BoM seems hardly to care if stations go bung. I remember your similar findings concerning gaps building to years in weather station records in the Perth catchment. Even though these records would be critical to evaluate run-off rates in an area with chronic water problems, the BoM just couldn’t be arsed to get out and fix the equipment.
Some change in the relationship between the BoM and their thousands of obervers?
AWS numbers have increased from under 300 in the mid 1990’s to ~700 now.
The crazy thing is that, from the point of view of measuring “climate change”, the BoM would be better off with a few dozen high-quality stations with complete records than with many hundreds with short records and gaps all over the place.