| Data and Documentation | Chronology and News | Program Status | Publications | Related Data and Resources | Contact Points |
ICOADS Web information page (Wednesday, 25-Jul-2012 18:07:50 UTC):
Twentieth Century Reanalysis version 2 (20CRv2) Feedback
This webpage presents examples of feedback information (Matsui and Compo 2011, hereafter MC11) available from version 2 of the Twentieth Century Reanalysis (20CRv2; Compo et al. 2011) for 1860-2008, as applied to ICOADS sea level pressure (SLP) data. These results are stratified temporally (Figs. 1-2, and 6) and by deck (Figs. 3-5). In addition, Figure 6a1 shows detailed feedback results for deck 780 (World Ocean Database), with Fig. 6b1 comparing SLP quality control ("trimming") information available from the latest ICOADS Release 2.5 (R2.5; Woodruff et al. 2011). Similarly, for other selected individual data sources, Figs. 6a2-6b2 show detailed feedback and trimming results for deck 702 (Norwegian Logbook Collection) [note: additional individual deck assessment are planned under Figs. 6 at a future date].
No ICOADS QC flags were used in the processing of input data for 20CR, but results in these figures are presented both: (a) filtered through the 20CR quality control indicator (QCi; assimilation feedback field 9 of MC11, which indicates simply whether a given observation was rejected or accepted by 20CR), and (b) unfiltered (note: these alternative results are revealed by activating the link at the end of each figure caption). The quantity plotted in colors in all these figures, ΔP, is the ensemble mean first guess pressure minus the modified observation pressure (assimilation feedback field 12 of MC11). In general terms, ΔP indicates how far off observed pressures were from first guess information used for 20CR, with red used to indicate large departures above or below (>9 hPa in absolute value).
Additionally we emphasize that the currently available R2.5 data were not actually used as input for the entire 20CRv2, thus Figs. 6a1 and 6b1 are not directly comparable (similarly for other individual deck comparison Fig. 6 pairs, showing data extending past 1951). Specifically, as discussed in Compo et al. (2011) the latest available version of ICOADS was used as the 20CRv2 was produced, such that ICOADS Release 2.4 was used for 1952-2008 and R2.5 for 1871-1951. Also some "Auxiliary" ICOADS (not yet blended into an official Release) sources were utilized by 20CR.
Year-to-year patterns (for July only) of QCi are provided in Fig. 7a, and of PF for comparison in Fig. 7b, through 1951. As shown in Fig. 7a, July 1944 had a relatively high percentage of observations rejected by QCi, so Fig. 7c compares the distribution of PF with that of QCi, for that single year-month.
Figure 1. January (1860-2008) filtered ensemble mean first guess pressure minus modified observation pressure (ΔP). Left axis: percentage of ΔP values (colors, hPa) per year-month; right axis: total number of ΔP values in thousands per year-month (line). (Vs. unfiltered by quality control indicator.)
Figure 2. July (1860-2008). (Otherwise as for Fig. 1.) Note the missing information in 1869, 1906 and 2002. (Vs. unfiltered by quality control indicator.)
Figure 3. For smaller decks (< ~25K): Combined January and July (1860-2008) filtered ensemble mean first guess pressure minus modified observation pressure (ΔP). Left axis: percentage of ΔP values (colors, hPa) per deck; right axis: total number of ΔP values in thousands per deck (line). Decks are defined in Tables D6a-D6c in the IMMA documentation. (Vs. unfiltered by quality control indicator.)
Figure 4. For larger decks (> ~25K): (Otherwise as for Fig. 3.) Note decks 702 (Norwegian Logbook Collection) and 780 (World Ocean Database). (Vs. unfiltered by quality control indicator.)
Figure 5. For the late 19th century: (Otherwise as for Fig. 3.) (Vs. unfiltered by quality control indicator.)
Figure 6a1. Deck 780 (World Ocean Database 2005 through 1951; earlier version after that date) combined January and July (1862-1996) filtered ensemble mean first guess pressure minus modified observation pressure (ΔP). Left axis: percentage of ΔP values (colors, hPa) per year-months; right axis: total number of ΔP values per year-months (line). (Vs. unfiltered by quality control indicator.)
Figure 6b1. Deck 780 (World Ocean Database 2005) combined January and July (1862-2004) sea level pressure trimming flag (PF). Left axis: percentage of PF values per year-months (colors); right axis: total number of PF values per year-months (line, logarithmic scale), where PF values are defined as follows:
| 1 | within 2.8σ limits |
| 2 | less than 2.8σ lower limit |
| 3 | greater than 2.8σ upper limit |
| 4 | less than 3.5σ lower limit |
| 5 | greater than 3.5σ upper limit |
| 6 | less than 4.5σ lower limit |
| 7 | greater than 4.5σ upper limit |
| C | limits missing (ocean/coastal box) |
Figure 6a2. Deck 702 (Norwegian Logbook Collection). (Otherwise as for Fig. 6a1.) (Vs. unfiltered by quality control indicator.)
Figure 6b2. Deck 702 (Norwegian Logbook Collection) combined January and July (1867-89). (Otherwise as for Fig. 6b1.)
Figure 7a. July (1871-1951) quality control indicator (QCi). Left axis: percentage of QCi values (colors) per year-month; right axis: total number of QCi values per year-month (line, logarithmic scale).
Figure 7b. July (1871-1951) sea level pressure trimming flag (PF). Left axis: percentage of PF values (colors) per year-month; right axis: total number of PF values per year-month (line, logarithmic scale).
Figure 7c. July 1944 quality control indicator (QCi) and sea level pressure trimming flag (PF). Left axis: percentage of QCi values (colors) per PF value; right axis: total number of flagged observations (i.e. matching QC and PF values; line, logarithmic scale).
References
McColl, C., N. Matsui, and G. Compo, 2011: HDF5 Format for International Surface Pressure Data Bank v10.11 (International_Surface_Pressure_Data_Bank_format_v10.11).
Compo, G.P., et al., 2011: The Twentieth Century Reanalysis Project. Q. J. R. Meteorol. Soc., 137, 1-28 (Jan. 2011 Part A) (doi:10.1002/qj.776).
Woodruff, S.D., S.J. Worley, S.J. Lubker, Z. Ji, J.E. Freeman, D.I. Berry, P. Brohan, E.C. Kent, R.W. Reynolds, S.R. Smith, and C. Wilkinson, 2011: ICOADS Release 2.5: Extensions and enhancements to the surface marine meteorological archive. Int. J. Climatol., 31, 951-967 (doi:10.1002/joc.2103).
[Documentation and Software][Links to additional]
U.S. National Oceanic and Atmospheric Administration hosts the icoads website privacy disclaimer
Document maintained by icoads@noaa.gov
Updated: Jul 25, 2012 18:07:50 UTC
http://icoads.noaa.gov/20th.html