Operational Oceanography Foresight

I found the foresight document you worked so hard on, very interesting. Here are a few more or less significant comments, to help you finalise the document.

In accordance with your suggestion, here are my comments on the draft document on Operational Oceanography foresight.

These comments were made by some of the IRD researchers/engineers. The document was circulated to all of the oceanographers at the IRD and evolved to take the comments into account. It thus reflects a substantial proportion of our community but not all of it. Nor is it an official message from the IRD which may be coming as well.

As promised, herewith below a few questions or suggestions. The more I dip into this OO foresight document, the more I find it well done and relevant! However, though I may be wrong (because there are some points I’m not clear about) it appears to be missing a few items, though I am not able to say how important they may be! (I’m deliberately trying to cover as much as I can, but will be relieved if it turns out that my fears are not justified).

The subjects I think are missing are:

“Climate Reprocessing”, by which I mean: “New efforts to establish *Climate quality* a posteriori based on observations which were not made for that purpose”! The idea is to promote/request a new episode similar to ESA’s Climate Change Initiative which was quite successful, among others, with respect to the MSL (in that it enabled convergence between the Envisat and Jason-1 time series). Would it be possible to envisage a similar approach for surface wind observed by scatterometers and altimeters? By improving forcing we should be able to improve the quality of Glorys type analyses, shouldn’t we ?   (or by increasing the already significant duration needed for a forecast to be realistic?).

2) Communication plan : the number of stakeholders concerned by progress of Operational Oceanography (OO) is increasing and they are scattered around the world., consequently I think there is a great risk of duplication and especially of people not using the OO (since they ignore the capabilities we’ve invested in), particularly as far the coastal domain is concerned ! (for details of the ONERC, BRGM meeting at the beginning of September, see below).   More specifically, I’m wondering whether the format of the GMMC days is still suitable for the many different stakeholders involved in the coastal domain. Shouldn’t the proponents of OO make a greater effort to ‘colonize’ other professional meetings? (whether for coastal risks or for the Exact Repeat Missions (ERMs), among others!).

3) Finite element models: These concern 1/36 or 1/108 of a degree, but never MEF/FES!  IF (which I’m not sure about), some significant stakeholders (LNHE?) use MEFs (which is not generally the case for the “offshore community”), is the issue of coupling between a ‘Cartesian grid’ model (that may not be the right term!) and an MEF covering part of the coastline being done properly?

Herewith a contribution from the whole of the Mercator R&D department to enable you to respond constructively, I hope, to the foresight document. We’re looking forward to meeting you soon to conclude discussions on this subject and to finalise this document, as it is important to us.

These comments are not a reply on behalf of our organisation as a whole to the version of the foresight document we were asked to reread: they are only reactions which are widespread in the coastal community. The foresight document does in fact reveal the lack of precision or mistakes due to a very immature fusion between the offshore and coastal communities. In its attempt to ignore the historical differences between these two communities, both in the report and in the foresight document, the document spreads some confusion.

We have written these comments in order to help improve the initial document which is a worthy basis on which to build: the comments aim to give a fair and honest description of the report and to propose a foresight document which better takes into account the achievements of operational offshore and coastal models and which best identifies the objectives for future operational oceanography covering the two domains and the means needed to achieve this.

Generally speaking, as a long-standing user of Mercator products and as a collaborator, I am very enthusiastic about this foresight document for several reasons.

I agree with the comment that “researchers should get involved in operational centres” help improve the system and products.

My specific comment is about Section 5.3:

A significant undertaking which I think is too mixed up with the others is that of ocean/atmosphere coupling which interfaces with forecasting. Such work might lead to an improvement in the quality of products in some regions (e.g. Eastern Edge-current systems). This research work could be dealt with separately since several communities would find it easier to identify with it and thus participate in the undertaking suggested in the foresight document.

From the seasonal forecasting perspective, forecasting systems run into the initialisation shock problem which may be attenuated by initialising the forecasting model with “corrected” products. The latter may be obtained in various ways, the simplest being the nudging method which consists in combining the coupled model solution and the Reanalysis. These initial conditions may also be produced both in coupled or semi-coupled mode. “semi-coupled’ refers here to partial coupling methods (e.g. coupling only on flows or anomalies) or using an ocean model with an atmospheric mixing-layer model (slab). This would no doubt require further development but it should enable a qualitative leap in improvement of forecasting products (this has been demonstrated in simplified systems) and could be considered as an effort to increase added-value significantly.

On the regional scale, in upwelling systems (which are generally very productive), the Reanalyses indicate significant bias which might be due to a poor estimation of the rotation of mean wind in ECMWF products on the regional scale (a bias which has also been observed in downscaled atmospheric products of the CORDEX type). These biases are increased by coupling in global coupled models (a recurrent issue from CMIP3 to CMIP5). This obstacle has to be overcome for downstream applications (bio coupling). One line of research currently investigated by the community interested in these regions (the IRD community is highly involved) consists in studying the fine-scale ocean-atmosphere interactions (regional to mesoscale) and developing the coupled ocean-atmosphere regional platform. These developments could help define strategies/methods for improving Mercator products in these specific regions.