For some reason best known to himself Anthony Watts has jumped on the “DMIGate” bandwagon started by Paul Homewood over on this side of the Atlantic a few days ago. In his latest article Mr. Watts quotes with approval the “Not A Lot Of People Know That” article which we have already covered in some depth.

Here yet again is one of my comments that recently ended up on the NALOPKT cutting room floor:

You will note that I was suggesting that Ron Clutz’s extremely selectively interpretation of some of Walt Meier’s academic papers left a lot to be desired. Particularly given the additional fuel added to the “skulduggery” fire by the Watts Up With That article it seemed sensible to phone up NASA and ask Walt for his views on the second hottest Arctic sea ice topic on the planet at the moment, according to Messrs. Clutz and Homewood at least. That is the relative merits of Multisensor Analyzed Sea Ice Extent (MASIE for short) versus the National Snow and Ice Data Center’s Sea Ice Index (SII for short) for determining Arctic sea ice trends.

According to the NSIDC:

MASIE sea ice products are developed from National Ice Center (NIC for short) data with support from the U.S. Navy and from NOAA. MASIE is hosted by NOAA@NSIDC.

whereas:

The Sea Ice Index provides a quick look at Arctic- and Antarctic-wide changes in sea ice. It is a source for consistent, up-to-date sea ice extent and concentration images and data values from November 1978 to the present.

As luck would have I managed to get through to Walt on my second attempt, and he graciously agreed to be interviewed at extremely short notice. He told me that whilst he now worked at NASA he used to be at the NSIDC, and still collaborated with them. Here are the edited highlights of his thoughts on “MASIE v SII”:

MASIE repackages data from the NIC, and incorporates an ice edge hand drawn by analysts working with whatever satellite data they have available at the time. It is an “operational” product designed to produce a “best effort” ice edge each day, based on whatever data may be available at the time.

Visual data is obviously not available in winter, and the ice edge is often obscured by clouds in summer. Synthetic Aperture Radar can “see in the dark” and through clouds, but suffers from different limitations. The whole of the Arctic isn’t covered every day for example. In addition, and unlike the SII, data from different satellite sensors is incorporated which means there are inevitably inconsistencies from day to day and from year to year. There is also an element of “human subjectivity” because different analysts are working with different sources of data from one day to the next. Since the quantity and quality of data varies the time series will not be consistent over time.

On the other hand the SII was designed to use a consistent methodology over a long period of time using a single type of sensor. 100% automatically processed passive microwave data is the “gold standard” when it comes to determining sea ice trends. It is subject to some biases and thus is not necessarily as accurate on a given day as MASIE. However, the biases are consistent over time, so the time series will be consistent over time. This means that year-to-year comparisons and trend estimates will be more accurate in the passive microwave data than in MASIE.

So there you have it. If you’re on the bridge of a vessel sailing in Arctic waters then MASIE is the right tool for the job. If on the other hand you’re sat in front of a computer trying to get the best estimate of trends in Arctic sea ice extent then the Sea Ice Index is what you’ll grab from your toolkit.

Having had a chance to examine the “evidence” of DMI “skulduggery” presented by Messrs. Watts and Homewood, Walt sent me a follow up email. Here is what it says:

Regarding DMI, the issue seems quite simple. The 30% plot is an older version that they stopped supporting as they transitioned to the 15% plot. I don’t know specifically why the 30% plot went awry, but there is generally automatic quality control done to make sure the final results are accurate and consistent. If such QC is not done, a lot of incorrect values can occur. I suspect that since the older version was no longer supported, the QC wasn’t being watched and something went wrong that they didn’t bother to fix (or maybe didn’t even notice) because the new 15% version is the official DMI output.

Not a lot of people know that, because Watts, Homewood et al. have developed the nasty habit of “snipping” comments to that effect as and when the mood takes them, which based on my own experience seems to be remarkably often in this day and age.