GFS seasonal bias

[showmethesnow]

The GFS with the changes this summer has yet to be tested during the winter. I was wondering with the differences in the seasonal weather pattern do the biases also change? I am curious because people keep commenting on this bias and that bias when it comes to the updated GFS and we may really not know yet what they are. Was wondering if a met or anyone else may know.

[GaWx]

According to the attached, the day 6 GFS has been overall pretty neutral with regard to the 500 mb height bias at day 6 for the northern hemisphere as a whole fwiw. Prior to the most recent major GFS mod., this graph consistently showed a cold bias for the N. Hemisphere going all of the way back to 5/15/2001. So, it has clearly warmed quite a bit with regard to the northern hemisphere as a result of this last major mod.. Of course, you can have a neutral hemispheric bias and a cold E US bias at the same time since other areas could have a warm bias to balance it out.

[dtk]

The GFS with the changes this summer has yet to be tested during the winter. I was wondering with the differences in the seasonal weather pattern do the biases also change? I am curious because people keep commenting on this bias and that bias when it comes to the updated GFS and we may really not know yet what they are. Was wondering if a met or anyone else may know.

Actually the new model was run retrospectively over the winter 2009/2010 season. The temperature/height bias is quite different than the old GFS (prior to upgrade). Similar to the plot GaWX showed, here are comparable 6 day time series for last winter [here, the red/e13d is the current GFS, where the black/gfs was the version that was operational at the time):

and

The nature of the changes that were put in this summer has made for substantial changes in the performance of the model (mostly good, some bad). This is NOT the old GFS anymore.......

[showmethesnow]

Thanks. I had forgotten that they were being run concurrently last winter. So basically we do have an idea of what biases to expect for this winter.

I didn't word it to well but I was curious to whether a bias is consistent throughout the year or whether it may change from the seasonal time of year (spring, summer, fall, winter). I have never seen mention of it in the years I have been on the board so I am guessing that it is probably consistent.

[dtk]

Thanks. I had forgotten that they were being run concurrently last winter. So basically we do have an idea of what biases to expect for this winter.

I didn't word it to well but I was curious to whether a bias is consistent throughout the year or whether it may change from the seasonal time of year (spring, summer, fall, winter). I have never seen mention of it in the years I have been on the board so I am guessing that it is probably consistent.

It depends on the level and variable (i.e. there is generally a warm bias in the mid levels now, but an apparent near-neutral to slight cold bias at lower levels). I haven't parsed through everything so I hate to overgeneralize...all I can say is to expect something different than what has been documented in past seasons.

[showmethesnow]

It depends on the level and variable (i.e. there is generally a warm bias in the mid levels now, but an apparent near-neutral to slight cold bias at lower levels). I haven't parsed through everything so I hate to overgeneralize...all I can say is to expect something different than what has been documented in past seasons.

Thanks for taking the time to answer me.