Guest Post by Willis Eschenbach
[UPDATE: Grrr … thanks to an alert commenter, I find that my previous numbers were out by a factor of ten. I love the web for that, my mistakes don’t last long. Doesn’t affect my conclusions, but it is most embarrassing. The netCDF file gave the units as metres, but then in the small print included a “scale factor” of .0001. Never seen that before. It’s taken a bit of time but I’ve corrected the text and graphics. Grrr again. Now that the numbers are correct, I could remove this notice. I won’t. When I’m wrong, I admit it.]
In a recent post entitled “Boy Child, Girl Child” I discussed my functional analysis of the El Nino-La Nina phenomenon. I noted that the Nino-Nina alteration functions as a giant pump moving water from the equatorial Pacific towards both poles. In that post I used this image from NOAA:
Figure 1. The Nino/Nina differences as shown by the TAU/TRITON moored buoys along the Equator. You are looking westward, across the equator in the Pacific Ocean, from a vantage point somewhere in the Andes Mountains in South America. The colored surfaces show TAO/TRITON ocean temperatures. The top surface is the sea-surface, from 8°N to 8°S and from 137°E to 95°W. The shape of the sea surface is determined by TAO/TRITON Dynamic Height data. The wide vertical surface is at 8°S and extends to 500 meters depth. The narrower vertical surface is at 95°W.
In Figure 1 you can see how the strength of the La Nina winds has actually hollowed out the surface of the ocean along the Equator … but unfortunately, there was no scale on that part of the drawing. How much, I wondered, did the sea level go up and down with the Nino-Nina alternation?
Pondering that, I thought I could use the sea level height to determine a minimum value for the amount of water actually pumped over the year-long pumping stroke from El Nino to La Nina.
So I needed a sea level height dataset. I turned to the trusty KNMI Climate Explorer. When you click on their “Daily Fields” link, down near the bottom of the resulting page is the Copernicus 1/4° satellite-based sea level data. I wanted the sea level gridded data in netCDF format. So I selected “sea level anomaly” among the Copernicus choices. And down at the very bottom of that type of page, KNMI generally offers a netCDF option. In this case, it says “C3S sea level anomalies is available as a netcdf file (size 8851.18 MB).
Yikes. Nine gigabytes. It’s a big dataset, but I’m blessed with gig-speed fiber optics to my desk, so it was about a ten-minute download.
I wanted to make a movie of the 1997/98 Nino/Nina pumping cycle. So from that big dataset, I extracted a subset from January 1, 1997, to December 31, 1998, and converted it into an animated GIF. I used every other day’s data to keep the size of the output down. I then used an online service to convert it to an MP4 movie file. That gives me (and you) control over the display that an animated GIF doesn’t have. It also reduces the size from 67 Mbytes for the GIF down to 14 Mbyes for the MP4, for faster loading.
As always, the result is both fascinating and educational. The peak of the El Nino, which is the intake stroke of the pump, is in November 1997. As noted in my previous post, the phenomenon always lasts ~ twelve months, so the end of the discharge stroke is in November 1998.
Figure 2. Movie of the 1997-1998 El Nino – La Nina pump cycle. White contour lines around bright red and dark blue areas show sea level elevations/depressions of +.2 and -.2 metres.
What did I notice in the movie? Well, lots of things. I noticed that I can see the eddies that extend from the southern tip of Africa clear around to below New Zealand’s South Island. I can see the eddies of the Gulf Stream in the western North Atlantic, as well as those of the corresponding Kuroshio Current in the western North Pacific. I can watch the comings and goings of the ice around Antarctica and in the Arctic.
I also note that in the Indian Ocean something is going on that looks very much like the Nino/Nina alteration in the Pacific … watts up with that?
What else? In November 1997 I can see the well-known effect of the El Nino depressing the ocean surface levels in the islands between Australia and the Equator.
I also notice that starting around March 8th, 1997, there’s a large bulge of seawater that develops around the Equator north of Australia. Curiously, although most equatorial Pacific water is moving westward, over the next six weeks this bulge moves as a wave eastward until it hits the coast of South America. When it hits, the El Nino immediately begins to form and extend out from the coast towards the mid-Pacific.
That eastward-moving wave goes about 9,000 miles (14,500 km) in 40 days or so. That means it’s moving at about 9 mph (14.5 kph). Curious.
In any case, as you may recall, I started this hejira because thought I could use the sea level height to determine a minimum value for the amount of water actually pumped over the year-long pumping stroke from El Nino to La Nina. So what I did was, I calculated the difference between the sea level heights at peak El Nino (November 1997) and peak La Nina (November 1998). Figures 3 to 5 show those results. To begin with, here is the average November 1997 sea height.
Figure 3. Sea surface height anomaly, November 1997.
Next, here’s the opposite end of the pumping cycle, in November 1998.
Figure 4. Sea level height anomaly, peak La Nina conditions, November 1998. Note the different legend range and colors.
Finally, here’s the difference between the two:
Figure 5. November 1998 (La Nina) sea level anomaly minus November 1997 (El Nino) sea level anomaly.
And this gives me the information I need to estimate the minimum amount of water pumped. I chose the rectangular area shown in red above as the main location of what is pumped. It goes from 5°N to 5°S, and from 170°W to 85°W. It is approximately the sum of the NINO34, NINO2, and NINO1 areas used to analyze the El Nino phenomenon.
For each gridcell in that area, I multiplied the surface height difference times the number of square metres in the gridcell. The sum of these is the total amount moved by the Nino-Nina pump cycle. This turns out to be about 3.6 trillion cubic metres of warm equatorial Pacific water pumped westwards and eventually polewards.
So … is 3.6 trillion cubic metres a lot or a little in a very big ocean? Well, the volume of water flowing in an oceanic current is measured in “sverdrups”. One sverdrup is a flow of a million cubic metres per second. Depending on where and when you measure it, the Gulf Stream flow ranges between about thirty to about a hundred and twenty sverdrups. And the AMOC, the Atlantic Meridional Overturning Current, runs at about twenty sverdrups.
And a flow of 3.6 trillion cubic metres per year works out to about 0.1 million cubic metres per second, which is 0.1 sverdrups.
As I said up near the top, I consider this a minimum amount moved. I say this because as the wind is hollowing out the area near the equator, water is tending to flow by gravity into that depression from all sides. This means that water is being pumped to the west even when there is no change in the sea surface level.
So the answer to my question that started this out is, over the year of its operation, at least a hundred thousand cubic metres every second is pumped westward and poleward by the Nino-Nina pump, for a total of at least 3.6 trillion cubic metres over the year.
What else? Well, by the end, the Nino-Nina cycle has exposed subsurface water that is about 4° cooler than the surface water at the start of the cycle … and it does so over about 10% of the entire tropical ocean.
And then, of course, there’s all the other stuff I learned along the way, which is its own reward.
Anyhow, that was my weekend. Did the research Friday and yesterday. Writing it up today, publish Monday morning … plus football. Gotta watch or at least listen to the games while I do other things. Like trimming away the redwood branches that have been hitting my overhead internet line, using my whizbang bolt-on adjustable ladder adapter for uneven ground that I designed and built a couple of days ago.
I started by U-bolting the four corners of the plywood in the center of the picture to a couple of rungs on the ladder. Then I adjustable bolted legs onto that. That lets me get 20′ (6m) up a redwood tree trunk here on my hillside in relative safety, whereupon I lash the top of the ladder to the tree with a ratchet strap. Works a treat, rock-solid.
Hey, do us 73-year-old geriatricats know how to party or what?
My best regards to everyone, and please, quote what you’re talking about. It saves endless misunderstanding. I can defend my own words. I can’t defend your interpretation of my words.