Thursday, September 17, 2015

New sea ice treatment in TempLS

TempLS is my global temperature anomaly calculator. It uses ERSST V4 for the sea surface component. ERSST unhelpfully records sea covered with ice as having the sea water freezing point of -1.8°C. Since this does not seem to be a good proxy for air temperature, I mark such points as missing data.

However, it seems -1.8 is not reliable - regions which are clearly frozen can report higher temperatures. Maybe there is an allowance for salinity variation. Anyway, to date I have been using ≤-1°C as the criterion for removal.

I now find that helps, but also doesn't work reliably. Even with the Arctic ice near peak, there seem to be a few deviant points reporting SST where there should be none. This creates a problem with my triangular mesh weighting in the Arctic. The idea is that land stations should be a better basis (than -1.8) for estimating nearby frozen seas. But these isolated spurious sea temperatures also become prominent in the mesh, and affect large areas, as if they were land.

So I am using a new criterion. For each SST location and month, I look at the record starting 1900, and count the occurrences of temperatures less than -1.5°C. If there are more than 10, I deem the location to be subject to intermittent freezing, and exclude it for that month throughout.

This will clearly exclude some valid data on the fringes of the ice. However, with some years frozen, others not, it is in any case hard to get an appropriate normal. In practice the decision about inclusion does not have dramatic effects for individual locations, since freezing leads to a zero anomaly. And because it is a lat/lon mesh, there is an artificially high node density anyway.

You may notice some small differences in the TempLS mesh results. The August average rose from 0.7°C to 0.703 °C.


  1. You could calculate the area covered by SST below a given temperature. If you have this area as a function of the temperature threshold, you could compare it to the sea ice area of the same time. Maybe there is a temperature threshold, where this area is close to the sea ice area. Good if this temperature is independent from month, but maybe it works also with a month dependent temperature.

    1. Yes, thanks, that is an interesting idea. I could compare the ERSST notion of sea ice with the NSIDC index, or some such. Maybe even compare pictures.

      I'm not sure of good results, though, because ERSST is really hard to pin down, because of the large area of what seems like it should be ice that has variable temperature. I made movies of SST like this with HiRes AVHRR data, to try to track the ice, and it kinda works, but with some variability, including patches of considerable warmth where ice might be expected. This may be the kind of data that is getting into ERSST.

  2. OT: does anybody know the precise boundaries of the SST area used in creating the AMO index? Abstracts and online definitions seem to just refer to it as the North Atlantic.

    1. I believe it's literally the North Atlantic. Look up the paper by Schlesinger (1993), which gives the first formal definition of it. (You can get the PDF off his researchgate page).

      Some people argue for subtracting off portions of it. You can find a discussion of that on the wiki page. Literally we're mixing apples and oranges here. We've no fundamental theory telling us how to compute the AMO (associated presumably with long-period thermohaline circulation), so other long period variability gets mixed in, including anthropogenic forcing, instrumental and methodological changes, and of course forced variations associated with other atmospheric-ocean oscillations (notably ENSO).

      The ocean boundaries are defined by the International Hydrographic Organization, in a document entitled "Limits of Oceans and Seas".

      There is a digital version of this maintained by Damien Fourcy, Link to his paper here. This paper discusses the issues of boundaries and standards in more detail.

      The shapefile for his digital map is freely available. You can download the shapefile here:

      There is a free R interface for shapefiles (Nick might be interested in pursuing this), but many people who do mapping use the ArcGIS program.

    2. I should note that Schlesinger uses singular spectral analysis to compute the AMO. I've not researched how other people calculate it, but I believe the calculation is more complex than just averaging over the SST of the North Atlantic.

    3. "Nick might be interested in pursuing this"
      Thanks Carrick, I did find this. I'll probably have to get into GIS some day. But for the land/sea issue, I'm thinking of just using the AVHRR SST files that I download. They give a grid of land/sea to 1/4° resolution, which should be enough.

    4. Hi Nick,

      I figured it would be a PITA to implement this for analysis, but I thought it might be useful for mapping purposes.

      By the way, there are SSA packages for R too. It's something I plan on exploring for my own data.

  3. I could not quite believe they were starting at the equator, but I guess they do. Given the persistent blue blob, it's somewhat surprising, to me anyway, the AMO is trending upward so far this year.

    1. JCH, it's worth noting that , starting with Schlesinger, they use singular spectral analysis (SSA) to separate the AMO pattern from the temperature field. In this case, you just need to bound the region where you observe the coherent behavior, rather than trying to carefully lasso it.

      The biggest problem you run into is spectral overlap between the AMO and other long period variability.