http://www.ncdc.noaa.gov/oa/climate/research/anomalies/anomalies.html
I have found nothing to indicate that the NOMADS ERSST data is based on the newest version, ERSST.v3, so this comparison should be based on the same data, ERSST.v2.
Figure 1 is the basic comparison between the two data sets, smoothed with a 37-month filter. The obvious difference is that the NOMADS ERSST data starts in 1854*, and the NCDC data begins in 1880. Other than that, there’s a minor apparent difference in the overall trends, with the NCDC data rising faster than the NOMADS ERSST data.
http://i28.tinypic.com/23h15wz.jpg
Figure 1
So I subtracted the NOMADS ERSST data from the NCDC version to create Figure 2 to see what differences there were. The minor differences could result from methods the two data sets use to address parts of the data: smoothing radius or blending, inclusion of more or less Arctic or Southern Oceans, or the inclusion of the Mediterranean Sea, any number of things. But look closer, there’s a step change visible near 1941 that raises the NCDC version considerably. If this is the bucket adjustment, why doesn’t it apply to both data sets?
http://i27.tinypic.com/t50p6h.jpg
Figure 2
In Figure 3, I’ve marked the approximate date of the change on the anomaly data graph.
http://i32.tinypic.com/oct2u.jpg
Figure 3
*A NOTE ABOUT THE DATA PRIOR TO 1880
A quote from the NCDC description of the ERSST data sets: “This monthly analysis begins January 1854, but because of sparse data the analyzed signal is heavily damped before 1880.”
http://www.cdc.noaa.gov/cdc/data.noaa.ersst.html
Damping minimizes periodic fluctuations, but it doesn’t radically alter the overall shape of an SST data curve.
SOURCE
Sea Surface Temperature Data is Smith and Reynolds Extended Reconstructed SST (ERSST.v2) available through the NOAA National Operational Model Archive & Distribution System (NOMADS).
http://nomads.ncdc.noaa.gov/#climatencdc
No comments:
Post a Comment