Thursday, April 30, 2015

UAH v6 trends

Roy Spencer has announced the beta version of V6 of the UAH lower troposphere temperature index. Sou has the story.

The result now look much more like RSS. The big things that made this happen seem to be that
  • UAH now uses a previously deprecated diurnal correction
  • UAH has reduced sensitivity to land surface emission, making it a more purely tropospheric measure (and more different from surface)

Incidentally, Dr Spencer's announcement is very informative on how the UAH sausage is made.

I was curious to see what difference it made to the back trend plots, as available here with the trendback button. It shows the trend from the various years on the x-axis to present. UAH is now similar to RSS, though not as low, and the pause is back. And the distinction between surface and troposphere indices is very clear. Here is the plot:


  1. Gavin sort of chewed me out when I said this, but to me the satellite series are not capable of any reasonable accuracy in estimating surface temperature unless there is a sort of balance between La Nina events and El Nino events. Since 2005, excluding the current El Nino, there is a pronounced imbalance between El Nino and La Nina events, and the satellite series are wandering around lost in the dark. It's still out of balance with the current event, but it has not yet been strong enough to be corrective.

  2. Upon reading Dr. Spencer’s announcement, I can only reflect on the fact that the challenges in producing a temperature record are at least equal, if not more difficult, with satellite data as compared to surface station data. From his annoucement I read that:

    1. There is data from sixteen satellites.
    2. There is instrument calibration and satellite orbital drift.
    3. The various satellites have different calibrations.
    4. There are frequency differences between satellites.
    5. Differences in time of observation.

    Concerning all this he states (emphasise mine): All data adjustments required to correct for these changes involve decisions regarding methodology, and different methodologies will lead to somewhat different results.

    And the GWPF wants to investigate the adjustments made to surface station temperature data??

  3. One of the differences between V6 and previous versions is the use of a single beam angle (as I understand it, this is a weighted average over all of the individual sensor beam angle). As Roy Spencer explains, this should improve the spatial resolution of the new product.

    One of the side effects of this new algorithm is that the mean height of the lower tropospheric measurement is elevated from about 3.5 km to 4.5 km.

    I'd expect this to have a bigger influence over land than ocean, and over regions with large vertical heat flux in the atmospheric boundary layer.

    It'll be very interesting to see a variant on Figure 4 where they show the differences in trends between version 6.0 and version 5.6.