is GOFS an improvement, otherwise what's the point?
Good question. Ditto for RASM-ESRL which seems to offer a third version of similar products (below). But that could be asked as well for UH and UB SMOS. Hamburg has coordinated salinity polarizability with Cryosat thickness; Bremen with another soil satellite SMAP. Maybe the two could get together and offer one optimal product?
http://ieeexplore.ieee.org/document/7730367/http://www.iup.uni-bremen.de/iuppage/psa/projects/SMOSice-project.phpsea temperatures at surface and depth matter
The air temperature at 2m can be quite cold relative to the sea water freezing point yet the re-analyzer temperature anomaly can still be a pronounced orange, especially for a 30 year base period that misses out on more recent Arctic Amplification (which is largely a fall and winter phenomenon).
Actually the temperatures of physical interest to the freeze season are those directly observable of the water and snow/ice surfaces themselves, not modeled meteorological 2m, though the three strongly influence each other.
Seductive computer graphics in many instances have gotten far ahead of actual data accuracy. Only a handful of products, such as UH SMOS, contain error analysis maps in their netCDF bundle. It is not rocket science, using 3rd party ImageJ 3D surface plugins, to drape say ice thickness over its error bump map, both as time series.
Note some open water north of the Bering Strait is still 3ºC above the freezing point. Given wind mixing (shown), thin dry air has a lot of work to do before stable frazil ice can form. Meanwhile upwelling net longwave cooling (provided by ESRL in Arctic11.gif, Arctic23.gif, and Arctic14.gif) is another consideration.
For all their shortcomings, we are probably better off using coupled radiative water-ice-snow-air-precip-cloud forecast models, which quantitatively integrate all the considerations, than intuiting off a single parameter such as a transient air mass.
https://www.esrl.noaa.gov/psd/forecasts/seaice/ navigate to Coupled --> Surface Fluxes
There are no active temperature gauges today in the Arctic Ocean itself, only a handful on the periphery and one daily sonde at Ny-Ålesund. This would be like producing a high resolution daily 2m temperature map of Europe using only station data from North Africa, Ireland and Finland, lol.
water vapor intrusions can have very significant impacts
Yes indeed, seems like last season had a number of notable and persistent events. Anybody recall the link to that very fine TPW web graphic? It showed counter-rotating water vapor trails sometimes rising up into the North Atlantic and beyond, bring warm vapor from the Caribbean.
Winter storms have been analyzed by L Boisvert and coworkers, including in several AGU2017 abstracts, notably:
GC43J-07: Increasing frequency and duration of Arctic winter warming events
http://onlinelibrary.wiley.com/doi/10.1002/2017GL073395/full open source
https://www.researchgate.net/profile/Linette_Boisvert/contributionsRM Graham et al
During the last three winter seasons, extreme warming events were observed over sea ice in the central Arctic Ocean. Each of these warming events were associated with temperatures close to or above 0°C, which lasted for between 1 and 3 days. Typically temperatures in the Arctic at this time of year are below −30°C. Here we study past temperature observations in the Arctic to investigate how common winter warming events are. We use temperature observations from expeditions such as Fram (1893–1896) and manned Soviet North Pole drifting ice stations from 1937 to 1991. These historic temperature records show that winter warming events have been observed over most of the Arctic Ocean.
Despite a thin network of observation sites, winter time temperatures above −5°C were directly observed approximately once every 3 years in the central Arctic Ocean between 1954 and 2010. Winter warming events are associated with storm systems originating in either the Atlantic or Pacific Oceans. Twice as many warming events originate from the Atlantic Ocean compared with the Pacific. These storms often penetrate across the North Pole. While observations of winter warming events date back to 1896, we find an increasing number of winter warming events in recent years.
Record low Arctic sea ice extents were observed during the last three winter seasons (March). During each of these winters, near-surface air temperatures close to 0°C were observed, in situ, over sea ice in the central Arctic. Recent media reports and scientific studies suggest that such winter warming events were unprecedented for the Arctic. Here we use in situ winter (December–March) temperature observations, such as those from Soviet North Pole drifting stations and ocean buoys, to determine how common Arctic winter warming events are.
Despite a limited observational network, temperatures exceeding −5°C were measured in situ during more than 30% of winters from 1954 to 2010, by either North Pole drifting stations or ocean buoys. Correlation coefficients between the atmospheric reanalysis, ERA-Interim, and these in-situ temperature records are shown to be on the order of 0.90.
This suggests that ERA-Interim is a suitable tool for studying Arctic winter warming events. Using the ERA-Interim record (1979–2016), we show that
the North Pole typically experiences 10 warming events (T2m > −10°C) per winter, compared with only five in the Pacific Central Arctic (PCA).
C21B-1119: Winter Arctic sea ice growth: current variability and projections for the coming decades
https://agu.confex.com/agu/fm17/meetingapp.cgi/Paper/258297C33C-1215: Rainy Days in the New Arctic: A Comprehensive Look at Precipitation from 8 Reanalysis
https://agu.confex.com/agu/fm17/meetingapp.cgi/Paper/253439