The data is free for use for
non-commercial applications, but please read the terms and conditions.
We request that users please register with the website here.
You do not have to register, but registering and providing information on the applications that the dataset is being put to will help us in
improving the science.
Audit Trail for HadTH
The process is documented in McCarthy et al. 2009 and summarised below with links to the data.
Source Data - HadTH0.0
The dataset is constructed from two datasets provided by the National Climatic Data Center in the U.S. These
are the Integrated Global Radiosonde Archive (IGRA) provided by NCDC to Hadley Centre at the
time of generation included data up to the year 2003. The dataset was further supplemented by an earlier product the Comprehensive Aerological Reference
Dataset (CARDS, Eskridge et al. 1995, Bull. Amer. Meteor. Soc., 76, 1759-1775) for the period 1963-2001 inclusive. As stated in
the manuscript this dataset is not suitable for assessing trends or variability in individual station records.
List of IGRA and CARD stations used in HadTH
HadTH0_0_day Normals - The unhomogenised 1981-2000 monthly climatologies from day launches.
HadTH0_0_night Normals - As above from night launches.
Correct Sampling biases - HadTH1.0
The next step in HadTH removed known sampling biases resulting from missing "dry" and missing "cold" observations early in the
A list of stations for which daily humidity data were adjusted where low humidity values
had been replaced with 19% relative humdity or dewpoint depression of -30degC.
Monthly means were also adjusted to account for missing humidity observations at low temperatures at many stations.
HadTH1_0_day Normals - The sampling bias corrected 1981-2000 monthly climatologies from day launches.
HadTH1_0_night Normals - As above from night launches.
First guess of known breakpoints - HadTH1.1
A set of adjustments are created for known metadata events such as change of instrumentation and included in Appendix C of
McCarthy et al. 2009.
Homogenisation - HadTH1.2
The dataset is then homogenised to produce:
Uncertainty - HadTH_Quarc
The homogenization method is conducted 100 times to produce an ensemble of reconstructions that can be used to quantify the
QUARC - 100 realisations of station adjustments.
Commercial and media enquiries
You can access the Met Office Customer Centre, any time of the day or night by phone, fax or e-mail. Trained staff will help you find the information or products that are right for you.
Contact the Met Office Customer Centre