Created by on - Updated on 10/07/2019 19:44

Several major collections of surface meteorological observations are

The Global Surface Network (GSN): The GSN provides a global reference network of land surface observation stations and mid-oceanic islands that can be used to detect and quantify aspects of climate change. The network has an approximate density of one station per 250,000 square kilometers. It is optimized for temperature observations and for suitability of stations for climate analysis.

The Integrated Surface Database (ISD): contains the full suite of meteorological variables (e.g., temperature, precipitation, winds) from land stations.

The International Comprehensive Ocean Atmosphere Data Set (ICOADS) 1662-present: contains the full suite of meteorological variables and also sea surface temperature from marine platforms such as ships and buoys.

The International Best Track Archive for Climate Stewardship (IBTrACS) 1851-present: contains tropical cyclone location and intensity reports from around the world.

The International Surface Pressure Databank (ISPD) 1722-2015: contains surface and sea level pressure observations from three components: land (largely from ISD and also 68 other national and international collections of pressure observations), marine (largely from ICOADS), and tropical cyclone pressure reports from IBTrACS. Discussion Page.

Annual Maps of the ISPD latest version station component can be browsed at

The International Surface Temperature Initiative is assembling station temperature and other meterological variables with an emphasis on traceability and provenance through development of a global land surface databank.

The Hadley Integrated Surface Dataset (HadISD) 1973-present: is subset of selected stations from the Integrated Surface Database (ISD) for selected variables being quality-controlled by the UK Met Office Hadley Centre.   Take part in a discussion of data quality control issues.

The European Climate Assessment & Dataset (ECA&D) ~1900-present: contains daily land station observations for >7500 stations throughout Europe and the Mediterranean. The dataset has 12 meteorological elements for which derived products such as trend maps and anomaly maps are created. Also a gridded version (E-OBS) is available. Discussion page.

The Southeast Asian Climate Assessment & Dataset (SACA&D): contains daily land station observations for  stations throughout Southeast Asia. The dataset has 10 meteorological elements for which derived products such as trend maps and anomaly maps are created.

discussion of Data Rescue to contribute to these and other collections are found on the Data Rescue pages


Historical Surface Weather Maps:

1878-1909 Sir Charles Todd Weather Folios, Australia and western Pacific

1871-2003 United States Daily Weather Maps




Tue, 02/11/2014 - 12:15

I am compiling a list of the documentation for the observation operator used by each reanalysis system to obtain surface pressure. For the 20CR, the H(x) to obtain surface pressure is documented in Compo et al., QJRMS, 2011, equations 5 to 7. Please add the appropriate reference for each system as a reply to this comment. I'll make a page with the results. best wishes, gil compo

macbenoy (not verified)

Tue, 11/19/2013 - 04:16

Gil, I know I cleared this query under private cover, but herewith for the record: The two maps you refer to came from different meteorologists.- the large coloured one done under the management of Charles Todd, Astronomer and Meteorological Observer of the Colony of South Australia, while the newsprint map was done by his equal, Ellery (I believe), of the Colony of New South Wales. There were 6 colonies until Australian Federation in 1901 and each ran its own Met Service. (the true date was about 1909 when the Federation process got around to fully integrating these services). I suppose the take on these synoptic charts is the difference of interpretations of the data. It's a topic that contemporary Australian forecasters can discuss endlessly between each of the (now) States - all taking their slant on the "true synoptic situation". However, I have to admit that the two maps you have identified here seemingly have major differences. Not being qualified to comment, I leave it to my betters to judge the two meteorologists involved. On the Todd map, you can actually read the datapoints that both men would have used to construct their charts, so the data is there for contemporary analysis. And, of course, we have digitised a large amount of this data and it now resides in the ISPD.

There is no difference between the B&W and colour maps of the same date - they are the same page. In our workflow, if a page contained printed text, it was scanned (slow) for later OCR. If it contained graphics (eg. synoptic chart) it was photographed (fast) to retain the colour copy..... thus there is redundancy in the images. I will send you under separate cover a pdf of the manifest and post it to at a later date.

Mac Benoy

Mac, The B&W and colour maps are visibly different. Do know what the original sources of the B&W maps are? Who drew those? Did the B&W maps have the same input stations, or did the coloured maps have more? Presumably, Todd and his team drew the colored maps. Was the B&W map their preliminary analysis made in time for newspaper publication while the coloured map is a post analysis (a reanalysis!) done as more observations became available? thanks, gil

Gil Compo (not verified)

Mon, 09/23/2013 - 17:35

Todd Project team, On many of the pages, there is more than one map of Australia and the contours are different. For example, at what is the difference in the source between the apparent newspaper printed black and white map and the hand-drawn map with red isobars? Also, could the Manifest at be made into a PDF? It currently is not searchable. thanks, gil compo University of Colorado/CIRES NOAA/Earth System Research Laboratory

Add new comment

This question is for testing whether or not you are a human visitor and to prevent automated spam submissions.