This site provides useful practical information related to global and national weather observing practices and instruments, including independent equipment reviews.
You will find much of the background in my book The Weather Observer’s Handbook (published by Cambridge University Press), details of which can also be found on this site, together with useful links and downloads.

Stephen Burt
Author, The Weather Observer’s Handbook

A nation obsessed with the weather? Yes, certainly, but which nation?


Tip of The Day
Months or years of data can be lost in an instant if held in a single file on one hard disk. An entire lifetime’s manuscript record could just as easily be lost forever in a house fire or burglary. Taking simple steps, including putting in place a multiple backup strategy, can hugely improve the chances that records (and instruments) will survive to be used by future researchers.
Before spending money on an automatic weather station (AWS) - consider firstly whether the site where the instruments will be used is suitable. There is little point in spending large sums on a sophisticated and flexible AWS if the location where it will be used is poorly exposed to the weather it seeks to measure. In general a budget AWS exposed in a good location will give more representative results than a poorly exposed top-of-the-range system. Worthwhile observations can be made with budget instruments in limited exposures, but a very sheltered site may not justify a significant investment in precision instruments, as the site characteristics may limit the accuracy and representativeness of the readings obtained.
Metadata is literally ‘data about data’. In the context of weather records, it is a description of the site and its surroundings, the instruments in use and any changes over time, information about observational databases and units used, and any other details about the measurements that may be relevant.
Recording raingauges can be easily and accurately calibrated by passing a known volume of water through the gauge, and comparing with the indicated measurement. ‘Out of the box’ errors for some AWS tipping-bucket raingauges of this type can exceed 20 per cent, so this is a vital test for all new instruments at first installation. Recording raingauges should not be adjusted merely to attempt exact agreement, or near-agreement, with a standard raingauge, because instrumental and exposure differences inevitably lead to slight variations in the amount of rainfall recorded.
It is advisable to check and test all sensor / datalogger / software and communications thoroughly, over a period of at least a few days, before permanent hardware installation or embarking on any long-term data collection.