Are Gulf Of Mexico Sea Surface Temperature Anomalies Near To Record Levels?

Bob Tisdale shows that once again, the devil is in the details.

Guest post by Bob Tisdale

OVERVIEW

This post illustrates Gulf of Mexico Sea Surface Temperature (SST) data and discusses the post by Jeff Masters at his www.wunderground.com blog in which he states that April 2011 Gulf of Mexico SST anomalies are “among the highest on record”, a claim that was then parroted by Joe Romm at Climate Progress. This post will illustrate that the current Sea Surface Temperature (SST) anomalies in the Gulf of Mexico are well below past weekly and monthly record high SST anomalies. It will also show that there is little to no long-term (80-year) trend in the sea surface temperatures of the Gulf of Mexico, a fact that holds true for the three long-term SST datasets (HADSST2, HADISST, and ERSST.v3b) used in global temperature anomaly products.

INTRODUCTION

I, like many, will occasionally drop by Climate Progress to see what Joe Romm is up to. The headline of the post “Masters: Midwest deluge enhanced by near-record Gulf of Mexico sea surface temperatures” caught my eye. Near record? Claims of “near record” and, in the body of the post, “among the highest on record”, give the impression of a long-term increase in Gulf of Mexico SST anomalies. They are clear attempts to imply that there are anthropogenic bases for the “Midwest deluge”. But I’ve studied Gulf of Mexico SST data, and there is little to no long-term trend in the SST anomalies there since the 1930s. So I continued to read Joe Romm’s post.

Romm writes, “Former hurricane-hunter Masters has a good analysis of how the ‘Midwest deluge [is] enhanced by near-record Gulf of Mexico sea surface temperatures…’” I’ll let you decide if it was a “good analysis”. And then Romm includes a quote from Jeff Masters’s post “Tornadoes, floods, and fires continue to pound U.S.”, which reads, “The deluge of rain that caused this flood found its genesis in a flow of warm, humid air coming from the Gulf of Mexico. Sea surface temperatures (SSTs) in the Gulf of Mexico are currently close to 1 °C above average. Only two Aprils since the 1800s (2002 and 1991) have had April SSTs more than 1 °C above average, so current SSTs are among the highest on record.” [Bold face by Romm.]

Jeff Masters includes a NOAA/NESDIS-based SST anomaly map as his Figure 5, with a caption that reads. “Figure 5. Departure of sea surface temperature from average for April 25, 2001. Image credit: NOAA/NESDIS.” His Figure 5 is included here as Figure 1.

Figure 1

The entire paragraph about Gulf of Mexico Sea Surface Temperature (SST) anomalies by Jeff Masters at his blog at Weather Underground reads:

“Midwest deluge enhanced by near-record Gulf of Mexico sea surface temperatures

“The deluge of rain that caused this flood found its genesis in a flow of warm, humid air coming from the Gulf of Mexico. Sea surface temperatures (SSTs )in the Gulf of Mexico are currently close to 1 °C above average. Only two Aprils since the 1800s (2002 and 1991) have had April SSTs more than 1 °C above average, so current SSTs are among the highest on record. These warm ocean temperatures helped set record high air temperatures in many locations in Texas yesterday, including Galveston (84°F, a tie with 1898), Del Rio (104°F, old record 103° in 1984), San Angelo (97°F, old record 96° in 1994). Record highs were also set on Monday in Baton Rouge and Shreveport in Louisiana, and in Austin, Mineral Wells, and Cotulla la Salle in Texas. Since this week’s storm brought plenty of cloud cover that kept temperatures from setting record highs in many locations, a more telling statistic of how warm this air mass was is the huge number of record high minimum temperature records that were set over the past two days. For example, the minimum temperature reached only 79°F in Brownsville, TX Monday morning, beating the previous record high minimum of 77°F set in 2006. In Texas, Austin, Houston, Port Arthur, Cotulla la Salle, Victoria, College Station, Victoria, Corpus Christi, McAllen, and Brownsville all set record high minimums on Monday, as did New Orleans, Lafayette, Monroe, Shreveport, and Alexandria in Louisiana, as well as Jackson and Tupelo in Mississippi. Since record amounts of water vapor can evaporate into air heated to record warm levels, it is not a surprise that incredible rains and unprecedented floods are resulting from this month’s near-record warm SSTs in the Gulf of Mexico.”

ARE GULF OF MEXICO SST ANOMALIES THIS MONTH AMONG THE HIGHEST ON RECORD?

I confirmed with Jeff Masters via WunderGround email that his approximation of a +1.0 deg C anomaly was based on the appearance of the NOAA/NESDIS-based map, not on the data itself. In other words, he looked at the scaling of the color-coded contour levels and estimated the Gulf of Mexico SST anomalies from them. He has compared an assumed DAILY Gulf of Mexico SST anomaly of 1.0 deg C to MONTHLY SST anomalies to arrive at the claim, “Only two Aprils since the 1800s (2002 and 1991) have had April SSTs more than 1 °C above average, so current SSTs are among the highest on record.” Was the Gulf of Mexico SST anomaly for that dataset on April 25th at approximately 1.0 deg C? Looks like it might be. But a one-day SST anomaly snapshot on April 25 does not represent the monthly average for April 2011. There is data available, so there’s no reason to attempt to read the SST anomalies from the temperature contour levels on a map. Since the month of April has not ended and monthly data is not available for it, why not simply use weekly data and average the data that’s available for the month? Why not? As we will see, it doesn’t provide an alarming answer.

The NOAA/NESDIS SST data used by Masters are prepared specifically for the NOAA Coral Reef Watch program. They exclude daytime satellite-based SST observations. This was discussed in my post A Note About SST Anomaly Maps. This is the only SST anomaly dataset that I’m aware of that excludes daytime data. Since the NOAA/NESDIS SST data excludes daytime SST data, and since the NOAA/NESDIS data is not readily available online in an easily usable format, we’ll use the other NOAA satellite-based SST dataset, Reynolds OI.v2, to determine the month-to-date values. Weekly and monthly observations for Reynolds OI.v2 data are available online through the NOAA NOMADS website:

http://nomad3.ncep.noaa.gov/cgi-bin/pdisp_sst.sh?lite=

Figure 2 shows the WEEKLY Gulf Of Mexico Sea Surface Temperature (SST) anomalies from January 3, 1990 (the start of the weekly Reynolds OI.v2 data centered on Wednesdays) through April 20, 2011. I’ve also included the average of the three weekly April 2011 SST anomaly values for the Gulf of Mexico (+0.79 deg C), shown as the red line. (The weekly observations are +0.68 for the week centered on April 6, 2011, and +0.91 for April 13, 2011, and +0.77 for April 20, 2011. Note that they dropped from the second to the third week.) Jeff Masters’s approximation for April is about 0.21 deg C higher than the 3-week average. Also, the weekly data clearly shows that Gulf of Mexico SST anomalies in excess of 0.79 deg C are a common occurrence, and that the elevated SST anomalies occur at differing times of the year. Note the spike in 2002. It occurred in April and May. The highest weekly April 2002 reading was +1.35 deg C, and it occurred during the week centered on April 24, 2002, but the highest April 2011 SST observation so far was only +0.91 deg C. The current April 2011 high is 0.44 deg C lower than the April 2002 high of +1.35 deg C. On a weekly basis, the current Gulf of Mexico SST anomalies are far from the past April record.

Figure 2

But Jeff Masters compared his assumed daily snapshot temperature to monthly data, not weekly data, so let’s examine the MONTHLY Reynolds OI.v2 data from its start year in November 1981. Refer to Figure 3. If we assume the April 2011 SST anomalies for the Gulf of Mexico will be near to the three-week average of 0.79 deg C, then the claim of “near-record warm SSTs in the Gulf of Mexico” is also a stretch. There is a 0.18 deg C difference between the current value and the record high of 0.97 deg C in April 2002.

Figure 3

Let’s also look at a graph of April SST anomalies using the Reynolds OI.v2 data, Figure 4, since Jeff Masters did specify the month of April. The values for 1982 and 2002 were the only times the current SST anomalies were exceeded. And again, there’s the 0.18 deg C difference between the present 3-week SST average anomaly and the record high in 2002. But note that the SST anomalies for the Gulf of Mexico do not exceed +1.0 in 1991 and 2002 as Jeff Masters had written. He also observed that this was the case “since the 1800s”, so he’s obviously comparing his assumed value to a long-term dataset, not a satellite-based dataset like the one in the map.

Figure 4

I tried the long-term HADISST and ERSST.v3b datasets and could not confirm Jeff Masters’s statements in his post, so I asked, and he advised the dataset he used was HADSST2. In summary, he’s compared monthly observations from a ship- and buoy-based SST dataset to an assumed daily snapshot of 1.0 deg C from a satellite-based SST dataset.

I considered stopping the post there. But let’s continue and take a more detailed look at the SST anomalies of the Gulf of Mexico. Let’s assume, similar to the assumption Jeff Masters has made, that the monthly HADSST2 data will be the same as the current 3-week anomaly of the Reynolds OI.v2 data for the Gulf of Mexico. Big assumption. Figure 5 shows the April HADSST2-based SST anomalies for the Gulf of Mexico. They confirm what Jeff Masters had written, which was, “Only two Aprils since the 1800s (2002 and 1991) have had April SSTs more than 1 °C above average…” But the SST anomalies for the month of April have not been reported yet. And the month-to-date SST anomalies for the Gulf of Mexico for April 2011 (based on the three weeks of April 2011 data) are far short of +1.0 deg C. So there’s no reason to compare to 1.0 deg C. Figure 5 also shows that SST anomalies for the Gulf of Mexico as high as (the 3-week average of) 0.79 deg C have occurred regularly, and they have occurred as far back as the 1920s.

Figure 5

NOTE: HADSST2 data is available as far back as 1850, but the source SST data for the Gulf of Mexico is sparse before 1900, leaving multiyear gaps in the data. Refer to Figure 6, which is the output graph created by the KNMI Climate Explorer. I’ve used January 1900 as the start month for the HADSST2 data for this reason.

Figure 6

THE MONTHLY HADSST2 DATA FOR THE GULF OF MEXICO IS REVEALING

Based on HADSST2 data, the monthly SST anomalies for the Gulf of Mexico have exceeded 0.79 deg C many times, even as far back as the early 1900s. Refer to Figure 7. Note also how flat the data has been since the 1930s.

Figure 7

If we start the data in January 1930, Figure 8, we can see the trend in Gulf of Mexico SST anomalies has been basically flat since then. It’s a noisy dataset, but it shows little long-term rise.

Figure 8

Therefore, if the April 2011 Gulf of Mexico SST anomaly should happen to equal or exceed Jeff Masters’s 1.0 deg C guesstimate, it is plainly not a consequence of any long-term (80-year) trend.

And if we smooth the data with a 13-month running-average filter, as shown in Figure 9, we can see that recent SST anomalies are approximately the same as they were in 1900.

Figure 9

THE RESULTS WITH HADISST AND ERSST.v3b DATASETS

HADSST2 is not the only long-term SST dataset that could be used to determine if the recent warming of the Gulf of Mexico is unusual. The Hadley Centre’s HADISST and NOAA’s ERSST.v3b are also available. Both are infilled, so there are no gaps in the data. That means the entire term of the datasets can be graphed easily. The differences between the ERSST.v3b, HADISST, and HADSST2 datasets are further discussed in the post An Overview Of Sea Surface Temperature Datasets Used In Global Temperature Products.

Figures 10, 11, 12, and 13 provide the same views of long-term data using HADISST data, from January 1870 to February 2011. (HADISST updates are delayed by a month.) As shown in Figures 10 and 11, there are fewer instances in the past when Gulf of Mexico SST anomalies exceeded the 3-week average of the Reynolds OI.v2 data for April 2011, with very few exceeding the assumed 1.0 deg C anomaly. HADISST might have been a better dataset for Jeff Masters to use for an alarmist post, except the HADISST Gulf of Mexico data also shows very little trend since 1930, as illustrated in Figure 12. And Figure 13 shows HADISST Gulf of Mexico SST anomalies that have been smoothed with a 13-month running-average filter. Note the multidecadal variability. I’ll have to compare that with the AMO. For those wondering about the steep decline at the end of the data, it starts in 2008 and should initially be a lagged response to the 2007/08 La Niña. Why didn’t the Gulf of Mexico SST anomalies rise is response to the 2009/10 El Niño? Dunno.

Figure 10

####################################

Figure 11

####################################

Figure 12

####################################

Figure 13

####################################

Similar graphs using the NOAA long-term SST dataset ERSST.v3b are included as Figures 14 through 17. ERSST.v3b runs from January 1854 to present. As you will see, it would have been the wrong dataset for an alarmist presentation. Figures 14 and 15 show that Gulf of Mexico SST anomalies were higher in late 1800s than they are today. Figure 16 shows that the linear trend for the Gulf of Mexico SST anomalies is negative since 1930. Nope, alarmists would not like the ERSST.v3b version of the Gulf of Mexico SST data. And the ERSST.v3b data for the Gulf of Mexico shows multidecadal variability quite clearly since 1900.

Figure 14

####################################

Figure 15

####################################

Figure 16

####################################

Figure 17

####################################

CLOSING

Jeff Masters’s claim that Gulf of Mexico SST anomalies are “among the highest on record”, which was repeated by Joe Romm, is contrived. It is based on a comparison of a monthly long-term SST dataset to a daily value assumed from the contour levels on a map. The assumed value of 1.0 deg C is 0.21 deg C higher than the three-week month-to-date SST anomalies for the Gulf of Mexico. Short-term satellite-based data show that the Gulf of Mexico SST anomalies are a noisy dataset, with the current anomalies well within the normal range of variability. Long-term SST anomaly data show that the trend of the Gulf of Mexico SST anomalies is flat or negative since 1930. In other words, over that past 80 years, there is no global warming signal in the Gulf of Mexico SST data.

SOURCES

The Reynolds OI.v2 SST data was downloaded from the NOAA NOMADS website:

http://nomad3.ncep.noaa.gov/cgi-bin/pdisp_sst.sh?lite=

All other data was downloaded from the KNMI Climate Explorer:

http://climexp.knmi.nl/selectfield_obs.cgi?someone@somewhere