Updated high‐resolution grids of monthly climatic observations – the CRU TS3.10 Dataset

TLDR: In this paper, an updated gridded climate dataset (referred to as CRU TS3.10) from monthly observations at meteorological stations across the world's land areas is presented.

Abstract: This paper describes the construction of an updated gridded climate dataset (referred to as CRU TS3.10) from monthly observations at meteorological stations across the world's land areas. Station anomalies (from 1961 to 1990 means) were interpolated into 0.5° latitude/longitude grid cells covering the global land surface (excluding Antarctica), and combined with an existing climatology to obtain absolute monthly values. The dataset includes six mostly independent climate variables (mean temperature, diurnal temperature range, precipitation, wet-day frequency, vapour pressure and cloud cover). Maximum and minimum temperatures have been arithmetically derived from these. Secondary variables (frost day frequency and potential evapotranspiration) have been estimated from the six primary variables using well-known formulae. Time series for hemispheric averages and 20 large sub-continental scale regions were calculated (for mean, maximum and minimum temperature and precipitation totals) and compared to a number of similar gridded products. The new dataset compares very favourably, with the major deviations mostly in regions and/or time periods with sparser observational data. CRU TS3.10 includes diagnostics associated with each interpolated value that indicates the number of stations used in the interpolation, allowing determination of the reliability of values in an objective way. This gridded product will be publicly available, including the input station series (http://www.cru.uea.ac.uk/ and http://badc.nerc.ac.uk/data/cru/). © 2013 Royal Meteorological Society

Full text

1
Updated high-resolution grids of monthly climatic observations
– the CRU TS3.10 Dataset
SUPPLEMENTARY INFORMATION
I. Harris, P. D. Jones, T.J. Osborn and D.H. Lister
Climatic Research Unit
School of Environmental Sciences
University of East Anglia
Norwich
NR4 7TJ
UK
1. Introduction
This document is a collection of supplementary information for the CRU TS3.10
paper. It contains:
• Web links to sources of data and information
• Detailed information about the updating of databases for each variable
• A brief analysis of the effect of using a wind climatology, rather than monthly
observations, in the calculation of PET.
• Supporting plots for the detailed trend analysis carried out with CRUTEM4
(for TMP) and GPCCv5 (for PRE).
2. Web Links
Organised by section.
Section 2.1
MCDW bulletins:
http://www1.ncdc.noaa.gov/pub/data/mcdw/ (eg, 'ssm0812.fin')
CLIMAT reports:
http://hadobs.metoffice.com/crutem3/data/station_updates/
Enquiries for the WWR CD-ROM can be made to wcdmp@wmo.int (according to
http://www.wmo.int/pages/prog/wcp/wcdmp/wwr/index_en.html). However, it is intended
that the CD material will be available online from the WDCA website after some corrections
have been made (personal communication by email, Christina Lief, NOAA; 11/05/10].
Section 2.2
Information document for BoM temperature series:
ftp://ftp.bom.gov.au/anon/home/ncc/www/change/HQdailyT/HQdailyT_info.pdf)

2
New Zealand temperature series:
http://www.niwa.co.nz/our-science/climate/news/all/nz-temp-record
Canadian homogenized series:
http://ec.gc.ca/dccha-ahccd/Default.asp?lang=En&n=B1F8423A-1
Real-time Canadian climate data:
http://climate.weatheroffice.gc.ca/climateData/canada_e.html
Section 2.3
Climate data for Iran: http://www.irimet.net/irimo/synoptic1.htm
Wet days database improvement:
Daily Precipitation stations: http://dss.ucar.edu/datasets/ds512.0/
Monthly Precipitation and RDY: http://dss.ucar.edu/datasets/ds570.0/
Section 2.4
Section 2.4.1
MCDW bulletins: http://www1.ncdc.noaa.gov/pub/data/mcdw/ (eg, 'ssm0812.fin')
Section 2.4.2
CLIMAT reports: http://hadobs.metoffice.com/crutem3/data/station_updates/
Section 2.4.3
BOM climate data: http://www.bom.gov.au/climate/data/
Section 3
NCAR Terrainbase elevation dataset: http://rda.ucar.edu/datasets/ds759.2/
Section 3.3
IDL: http://www.ittvis.com/ProductServices/IDL.aspx
Section 4.1
UDEL Temperature Data:
http://climate.geog.udel.edu/~climate/html_pages/README.ghcn_ts2.html
GPCC Precipitation Data: http://gpcc.dwd.de
Section 4.3.1
CRUTEM3 and CRUTEM4 Data: http://www.cru.uea.ac.uk/cru/data/temperature/

3
3. Detailed Information: Considerations for Updating Specific Variables
Mean temperature (TMP):
The CRU monthly global TMP database (Jones and Moberg, 2003) holds more than 5,000
station series. These are routinely updated through the use of CLIMAT and MCDW, and
with WWR when those decadal publications become available. For the current study, the
most recent (1991-2000) WWR decade became available from NCDC in 2004. Whilst the
monthly CLIMAT and MCDW should update most current WMO reporting series in the
archive, this is not met in practice. WWR data series offer the potential to infill missing
values and highlight possible quality-control (QC) problems in the existing real-time
CLIMAT/MCDW-updated series.
The merger of the WWR 1991-2000 series with the existing TMP archive did infill some
missing values. The percentage of missing values in the CRU dataset was reduced from 38%
to 34% during the period 1991-2000. In addition, a comparison between our existing TMP
database and the WWR data identified a number of potential quality-control problems in
either or both datasets. These shortcomings were identified and, where possible, rectified in
our TMP database.
Mean-monthly maximum and minimum temperature (TMX/TMN):
The CRU TMX and TMN archives (updated from New et al., 1999, 2000, by MJ05) were the
basis for the extended/updated TMX and TMN archives. Before the current study, these
archives held about 7300 series. The main sources of new data were:
 NOAA/NCDC (archives provided by Russell Vose, NCDC)
 CLIMAT – CLIMAT messages have included TMN and TMX data since November
1994
 Canadian archive of homogenized series (Vincent, 1998 and Vincent and Gullet,
1999)
 Australian TMX and TMN archive for the period 2000-2008 (David Jones, BoM,
pers. comm. 1 Feb 2006)
The NOAA/NCDC archives do not identify all stations with WMO Station Identifiers as
generally used elsewhere in the CRU archives. A matching exercise was undertaken using
latitudes, longitudes, station names and elevation. However, such an automated approach,
which does not include the matching of overlapping data values, cannot place all series into
“known” or “unknown” categories. Of the approximately 10600 station TMX/TMN series

4
received, about 3200 remained unidentified. These series may or may not be new to the
archive but they have been retained for the later gridding operation but have a lower priority
(the priority system for data use in gridding is discussed in Section 3).
The result of the merger between the MJ05 database and the NOAA/NCDC database was
then merged with the other three source files listed above. After the final merger operation,
the TMX and TMN archives each held over 14000 series (including series without WMO
Station Identifiers).
Monthly-total precipitation (PRE)
Before the current study, the CRU Precipitation archive (updated from Hulme, 1994, by
MJ05) held around 12400 station series. This was, and is, updated periodically from
CLIMAT and MCDW sources. In addition, contacts and collaborations (as above, for
temperature) provided new series and updates to existing series. The principal additional
sources used for precipitation were:
 A significant collection of monthly precipitation series - supplied by NCDC.
 WWR 1991-2000
 CD-ROM from the Association of Southeast Asian Nations (ASEAN): The ASEAN
Compendium of Climate Statistics.
 Climate data for Iran – a link to this data is given in Section 1 of the Supplementary
Information.
Many of the NCDC data series were not identified by WMO Station Identifiers (in a similar
manner to that described above for TMX/TMN). Thus a matching exercise was necessary,
which matched only 2260 out of the approximately 3800 station series received. As with the
TMX and TMN series, the remaining 1540 series may or may not be new to the archive, and
they have been retained for the gridding operation but given a lower priority.
The result of the merger between the CRU archive and the NCDC series was then merged
with all of the other (source) inputs. As with the TMP/TMX/TMN mergers, comparison of
overlapping values from the different sources led to the discovery of individual erroneous
values, and appropriate alterations were made. At the end of all merger operations, the
precipitation archive held just under 15000 series.

5
Monthly counts of rain-days (WET, threshold ≈ 0.1mm)
In the following explanation, 'RD0' is used to refer to monthly wet day counts with a
threshold of usually ≥0.1mm but sometimes up to ≥0.4mm (depending on the reporting
station), and 'RDY' refers to monthly rain day counts with a threshold of ≥1mm. The CRU TS
'WET' variable is the monthly count of wet days with ≥0.1mm.
The existing WET database ran to 1989 and had limited coverage, especially before 1970.
Since incoming data from MCDW and CLIMAT are 'RDY' observations, conversion factors
from RDY to RD0 were derived as follows. Daily station precipitation measurements were
obtained from UCAR (dataset ds512.0), and were processed to produce matched sets of RD0
and RDY counts. Only whole months of daily data were used, leading to temporally-
intermittent coverage. Average monthly conversion factors (from RDY to RD0) were then
derived as the ratio of RD0/RDY, and averaged into 5° latitude bands. Coverage at high
latitudes was poor, and values from adjacent bands were repeated if there were zero or only a
single station in latitude band. The station mean and latitude band mean factors for January
and July are shown in Figure S1. Monthly PRE and RDY were also obtained from UCAR
(dataset ds570.0). The RDY station data were converted to RD0 using the above factors, and
station normals for RD0 produced. The PRE station data, where matching stations were found
in the RD0 normals, were then converted to RD0 using equations (4) and (5) (with x = 0.45)
from New et al (2000). The resulting RD0 station database, covering the period 1961-2004,
was then merged with the existing WET database to form a 'final' database. The RDY-to-
RD0 banded conversion factors are also used in processing RDY from MCDW and CLIMAT
bulletins.

Frequently asked questions

Q1. Why is the PET calculated using a 1961-1990 climatology?

Because of the insufficient number of monthly wind observations available, thecalculation of PET is made using a 1961-1990 wind speed climatology. 

Q2. What were the shortcomings of the merger?

After the final merger operation, the TMX and TMN archives each held over 14000 series (including series without WMO Station Identifiers). 

Q3. What are the main sources of new data?

The main sources of new data were: NOAA/NCDC (archives provided by Russell Vose, NCDC) CLIMAT – CLIMAT messages have included TMN and TMX data since November1994 Canadian archive of homogenized series (Vincent, 1998 and Vincent and Gullet,1999) Australian TMX and TMN archive for the period 2000-2008 (David Jones, BoM,pers. comm. 

Q4. What is the iff of the wind speed observations?

Since the availability of wind speed observations is too sparse to allow their gridding, and use in the PET calculation, this exercise is for information only.m ax. c limd iff=0. 

Q5. What was the conversion factor used to convert the PRE station data to RD0?

The PRE station data, where matching stations were found in the RD0 normals, were then converted to RD0 using equations (4) and (5) (with x = 0.45) from New et al (2000). 

Q6. What is the potential of the WWR data series?

WWR data series offer the potential to infill missing values and highlight possible quality-control (QC) problems in the existing real-time CLIMAT/MCDW-updated series. 

Q7. How many stations are in the CRU monthly global TMP database?

Section 4.3.1 CRUTEM3 and CRUTEM4 Data: http://www.cru.uea.ac.uk/cru/data/temperature/The CRU monthly global TMP database (Jones and Moberg, 2003) holds more than 5,000 station series. 

Q8. What was the resulting RD0 station database?

The resulting RD0 station database, covering the period 1961-2004, was then merged with the existing WET database to form a 'final' database.