HYCOM GOFS 3.1 Re-Analysis Issues WA Coast

[Gala, John (ECY)]

Hi,

 

My name is John Gala and I work for the Washington Department of Ecology. We have been using GOFS 3.1 Re-Analysis data near the Washington and Oregon coast (bounding box [43 to 52, -129.52 to -123]) and noticed some issues with the year 2008 that we have not seen in our other years of interest (2000, 2006, and 2014).  Most notably we are seeing extremely high temperatures (50 degrees C) in waters of >200 meters and also at times extremely low values (-12.5 C). There is also similar behavior with salinity. I noticed on HYCOM.org there is a disclaimer for GOFS 3.1 Re-Analysis about this problem and was wondering if there are any other years that you know of that may have been impacted? Additionally, are there any plans in the future to fix this issue?

 

Thanks,

 

John

[Alan Wallcraft]

If these occur at the bottom (i.e. the deepest value that is not a data void) then this is almost certainly fixed in the source code by  Added the macro MASSLESS_1MM

I'm not sure if this is set in Navy's ESPC system that will replace GOFS 3.1.

If strange values occur that are unconnected to the bottom, then provide a date when this happens and I'll look into it.

Alan.

[Gala, John (ECY)]

Hi Allan,

 

There are two different issues that I noticed related to the bottom. For specific point locations of interest to us HYCOM layer 23 has always been the deepest non void layer. In the 2008 output however, layers below 23 are sometimes active and when they are they have extremely high or low temperature and salinity. The second issue is that for the 23 layers in our locations of interest that we expect to be active, the bottom layer (layer 23) is at times significantly warmer than the surface. For example the output for 2008-02-22T03:00:00 shows layer 23 to have a max temperature of 52 degrees C and a minimum temperature of -12.5 degrees C.  I have attached the output for 02/22/2008 (data_2008) retrieved from HYCOM.org netcdf subset. Additionally, I attached plots that show the WA coast in layer 23 on a normal day (01/01/2008) and on 02/22/2008 (in this figure it even shows temperature values on land). Thanks so much for your help!

 

Best,

 

John

 

From: Alan Wallcraft <alan.wa...@hycom.org>
Sent: Wednesday, May 22, 2024 7:35 AM
To: HYCOM.org Forum <fo...@hycom.org>
Cc: Gala, John (ECY) <jog...@ECY.WA.GOV>; Ahmed, Anise (ECY) <AAH...@ECY.WA.GOV>; Figueroa-Kaminsky, C.(ECY) <cfi...@ECY.WA.GOV>
Subject: Re: HYCOM GOFS 3.1 Re-Analysis Issues WA Coast

 

External Email

[Alan Wallcraft]

The file hycom_GLBv0.08_535_2008022112_t015.nc is corrupted (on our server and on the Navy server) and should not be used.  Note that t018 and t021 are not available for that day.

Please let us know of any similar corrupted days with values over land and bellow the bottom.

If level 23 looks OK except at a few points then it is likely the "MASSLESS_1MM" issue and if so this could occur in the same locations at other times in GOFS 3.1 Reanalysis and Analysis.

Alan.