Thursday, August 17, 2023

Practical Use of ASCAT Winds in Weather Routing

ASCAT winds are satellite scatterometer measurements of wind speed and direction on the surface of the ocean and other large bodies of water, normalized to a 10 m height. The data resolution is 25 km, which is comparable to the resolution of typical global weather models available to the public such as GFS (27.8 km) and ECMWF (44 km). New data are available typically four times a day, from various combinations of two satellites, Metop-B and Metop-C, tracking N-NE (ascending) or S-SW (descending), bringing us data from either the port or starboard side of their split data swaths. For background in ASCAT, see  There is a video at the end here demonstrating the use of the data.

Before presenting the specifics of how to use this data, let me back up a moment and put this in perspective. Wind speed and direction are the key factors in marine navigation, in large part because the wind makes the waves, which can be a hazard to any size vessel. For sailors, it is even more primary because wind is their engine. Thus this is the most fundamental information available. It is like having the ocean covered with thousands of buoys measuring wind speed and direction.  But unlike isolated real ocean buoys who give us data every 10 min in some cases, hourly at worst, the ASCAT data covers large swaths of the ocean but only give us data 3 or 4 times a day. 

In a sense, the ASCAT data have done a major job for us even if we do not look at it, because it is a crucial input to the global weather models whose forecasts we must rely on for routing. But even though the models have assimilated the latest ASCAT data, the model forecasts are not guaranteed to be correct. It is specifically not a goal of the models to reproduce the ASCAT and buoy observations. They have a broader goal to produce the best overall forecast at many levels of the atmosphere. In short, the model forecasts may in fact not be correct in some circumstances, which is why we must compare several models to decide which is likely the best.

And indeed, circling back to check a model forecast with actual ASCAT measurements at a specific place and time we care about is one of the primary reasons for us to access the ASCAT data ourselves. If one model forecast agrees with the ASCAT measurements and another does not, it is most likely the better one to use for our routing computations

Another reason to continually monitor the ASCAT wind measurements for our intended route is the occasional observation of localized anomalous flow. We might spot a notable hole in the wind or a shear line that is not apparent in the global model forecast. In these cases, we have to realize that the ASCAT are real data. That is what the wind was doing at that time, regardless of what the model might imply or not show, and we need to route with that in mind. 

The other aspect of "perspective" is to recognize that even though these ASCAT wind measurements are indeed the most fundamental data we care about, the use of this data, which takes several easy, but non-standard procedures,  is definitely in the realm of "going the extra mile" to learn the very most we can  about the wind ahead of us. For a racing sailor, this is standard operating procedure, but when cruising it would be called on less often, unless we are negotiating a dangerous wind pattern or, more likely, some light air pattern in which we are just  looking for enough wind to get there. 

We have presented this perspective in the past, and in earlier editions of our text Modern Marine Weather, but then after outlining the basic guides to getting the data by standard procedures, we moved on. Based on discussions with practicing navigators, however, it seems that we needed a more specific procedure for accessing this crucial data, and that is what we have created.

We have semi-empirically compiled graphic indexes of the available data and pass times for the typical cruising and racing waters around North America and Europe, presented below, as well as ways to automatically access the latest available data of interest. (Note that we are not covering here the even more convenient means of obtaining this data in grib format, which can be achieved with LuckGrib or Expedition. Users of those two fine apps, might still find some benefit from the timing structure we present here.)

We assigned the names to these regions; they are not official.

Each of these named regions provide a potential of 4 data opportunities in each 24 hr period, and the only way to see which might be latest data available is to download all 4 images. The file size varies from 20 to 40 kb, depending on how much data it includes. The UTC times shown (±1 hr) tell us satellite passage times of the 4 opportunities for new data. They occur in pairs, 13 hours apart, where the two passes of each pair are about an hour apart.

For example, in Biscay, we have data passes at 10:30 and 21:30, ± 1 hr. If I go to the ASCAT web site now and look at what passes took place yesterday, I get the files shown below. 

In the top we see the two passes at 21:30 ± 1h and on the bottom the two at 10:30 ± 1hr.  These are the actual live transit times of the satellites, representing the valid times of the wind measurements. On other days, these times will be different, but remain within these windows. 

But we do not get this data instantly. It takes 2 to 3 hr to process the data, so for the Biscay region, we would only look for new data at about 00:30 and 1430. Note that this delay time or latency is about the same as it is with the model forecasts, which also take 2 to 3 hr to process for our consumption in grib format. 

To clarify the concept that there will be new data about 4 times every 24hr: consider the example of San Francisco with valid times of 0525 and 1825 UTC, ± 1 hr, noting that it takes about 2 hr to process the data.

Thus while underway or planning a route across these waters, we would go to Google Earth (or wherever we are looking at the data) at about 0725 or later UTC and download all four passes given. We do not know ahead of time which of the four will bring the new data but most likely two or them will have a swath of new data that will be valid at 0525 ± 1hr. Then again at 2025 or later we would again download all four of the passes and among those will likely be two new data swaths valid at 1825 ± 1hr.

That is in effect a cookbook approach to the data, using our indexes as guides for when to look. We will not get new data between these two periods. We then have to correlate the valid times with the model times and forecasts at hand.

There are several ways to get the data underway. You can request the images from Saildocs or you can use the custom KML files we made to use in either qtVlm or the desktop version of Google Earth, which is a free download for Mac or PC. The KLM files can also be adapted for use in OpenCPN. For the last two methods at sea, we need to link our internet connection to a satphone or Iridium Go type device.  With those procedures the links can be stored in the apps and accessed with a button click.

To request the files from Saildocs, use this request structure for Biscay (green number from our index), meaning send this text in the body to (colors added to show what changes for each file, but using this method, you just need to change the 133 for Biscay to say 86 for Bermuda.


This will get you the 4 images of the data that you can then analyze manually from the graphics alone. They have a Lat-Lon grid as shown above but will not be georeferenced into a nav program. But the  main point is you can get the data that easily and our indexes show what to ask for and when.

Our ASCAT page links to articles and demos with details, plus has a link to get all of the KML files and the graphic indexes. We will add more examples as soon as possible, and in particular the process of comparing with model forecasts.

Viewing ASCAT winds in Google Earth desktop version (using the Starpath indexes above.)


Aaron Williams said...

How can I get the Universal Surface Analysis and Marine Observations Buoy data for Google Earth, the zip file does not have the two files any more?

David Burch said...

The ones we showed were actually custom files we made for students of our Marine Weather Course ( ). They were based on versions originally online but the online originals were broken. We reported this to the OPC and NDBC and maybe they are fixed now. On the OPC site, it is top link on the GIS data ( ) and it looks like the new NDBC has links on its home page ( ). Those are not the ones we started with.

Both have to be tested. Originally, the OPC version showed a map with lat wrong by 20ยบ. My guess is they are both fixed now, and indeed great resources.