Guest post by Bob Tisdale
SkepticalScience recently published a post titled Mercury rising: Greater L.A. to heat up an average 4 to 5 degrees by mid-century. It’s a cross post of a UCLA press release with the same title. It struck me odd, because we recently showed that Western North American land surface temperatures have declined in recent years. Refer to Figure 6 in the post IPCC Models vs Observations – Land Surface Temperature Anomalies for the Last 30 Years on a Regional Basis.
The press release is based on the Hall et al (2012) climate model study Mid-Century Warming in the Los Angeles Region published at the website C-Change.La.
So, does a 4 to 5 deg F rise in Greater Los Angeles land surface temperature anomalies by 2050 sound realistic? That equates to a rise of 2.5 to 2.8 deg C. Considering that, based on a preliminary look at the data, greater Los Angeles land surface temperature anomalies have been cooling for the 3-plus decades, it seems to be a real stretch of the imagination.
The GHCN-CAMS land surface temperature dataset comes in a number of resolutions, including 0.5 degree latitude and longitude. Lucky for us, it’s available through the KNMI Climate Explorer. And that means we can capture data for some reasonably small geographical areas. The GHCN-CAMS land surface temperature dataset was presented in the Fan and Dool (2007) paper A global monthly land surface air temperature analysis for 1948-present. It will allow us to get an idea of what Greater Los Angeles surface temperatures have been doing since 1948.
The UCLA press release states (my bold face):
Some of the smallest changes predicted, yet still nearing a 4-degree increase, are in Oxnard (3.68 degrees), Venice (3.70), Santa Barbara (3.73), Santa Monica (3.74), San Pedro (3.78), Torrance (3.80), Long Beach (3.82) and Santa Ana (3.85). Among the highest predicted increases are Wrightwood (5.37), Big Bear Lake (5.23), Palm Springs (5.15), Palmdale (4.92), Lancaster (4.87), Bakersfield (4.48) and Santa Clarita (4.44). Table 2 in the study calls out 27 distinct locations, such as downtown Los Angeles (3.92), San Fernando (4.19), Woodland Hills (4.26), Eagle Rock (3.98), Pasadena (4.05), Pomona (4.09), Glendale (3.99) and Riverside (4.23).
So they’ve made predictions for an area larger than the City or County of Los Angeles. For the sake of discussion, let’s say it represents the area bordered by the coordinates of 33N-35.5N, 121W-117W. See Figure 1. Those coordinates fit with the 0.5 degree grids. And we’ll call that dataset Santa Barbara-Ventura-Los Angeles-Orange Counties.
Figure 1
Figure 2 presents a time-series graph of the Santa Barbara-Ventura-Los Angeles-Orange Counties land surface temperature anomalies since January 1948. The data has a linear trend of 0.177 deg C/decade. In order for the land surface temperatures for that dataset to rise 2.5 deg C by 2050, the linear trend of the data has to change drastically to about 0.667 deg C/decade from June 2012 through December 2050.
Figure 2
In Figure 3, I’ve smoothed the Santa Barbara-Ventura-Los Angeles-Orange Counties land surface temperature anomalies with a 13-month running-average filter to reduce some of the variability. What caught my eye was the shift in 1976 that coincides with the Pacific Climate Shift. Curiously, it appears the dataset has been cooling since that shift.
Figure 3
Let’s take a look at the linear trends before and after the 1976 Pacific Climate Shift. We’ll switch back to the “raw” data. Before the climate shift, January 1948 to December 1975, the Santa Barbara-Ventura-Los Angeles-Orange Counties land surface temperature anomalies rose at a rate of only 0.108 deg C/decade, and after, from January 1977 to May 2012, they’ve cooled at a rate of -0.082 deg C/decade.
Figure 4
In Figure 5, I’ve added the projection of about 2.5 deg warming by 2050 to the graph to show how unrealistic that projection looks, especially when we consider that surface temperatures for the Santa Barbara-Ventura-Los Angeles-Orange Counties data have been dropping for 3+ decades.
Figure 5
Maybe I looked at too large an area. Let’s take a look at the data for the coordinates of 33.5N-34.5N, 118.5W-117.5W. See Figure 6 for the location. We call that dataset “Los Angeles Plus.”
Figure 6
As shown in Figure 7, the pre-1976 warming rate for the “Los Angeles Plus” land surface temperature anomalies is greater than the larger dataset, at about 0.33 deg C/decade. But the post-1976 trend is still negative at -0.074 deg C/decade.
Figure 7
One last try: Let’s decrease the area of the data again, Figure 8, this time looking at the land surface temperature anomalies for the coordinates of 34N-34.5N, 119W-118W. That captures Malibu, the Valley and much of the City of Los Angeles.
Figure 8
Doesn’t help. As shown in Figure 9, the trend after the 1976 Pacific Climate Shift for the Malibu-The Valley-Los Angeles land surface temperatures is negative at -0.086 deg C/decade.
Figure 9
CLOSING
Based on this quick look at land surface temperature data for the Greater Los Angeles area, the Hall et al (2012) study referred to in the UCLA press release appears to have no basis in reality.
SOURCE
The GHCN-CAMS land surface temperature data presented in this post is available through the KNMI Climate Explorer.
As a process improvement specialist, I have spent a lot of time looking at what we call run charts. Basically a plot of some single metric against time, which is what we have here. So while I have zero expertise in climate or weather, as an American Society for Quality Certified Quality Engineer, I have considerable knowledge of how to look at and interpet a run chart.
First, to draw a linear trend line for the entire run of data since 1948 is essentially meaningless. When one end of the data has an increasingly bad fit to a line, either there is more than one trend, or the data is not well represented by linear trends. econd, looking at figure 3, I think one could make a case for the temperature to have two “stable” albeit very noisy states. One centered at -0.5C, from <1948-1958, 1961-1976, and 2009-present. The other, centered at about +0.5C, from 1959-1960 and 1977-2008. The clearest shift is 1976, but if current temps continue, what we have is a shift back. It is a VERY noisy signal, but it really looks more like several shifts, than any kind of "trend" to me.
Is that UHI effect?
Just for informational reference. The two VERY RED blocks over the ocean represent an area of the coast that consistently has the highest winds of any area in Southern California. The winds get so high in that little “channel” that yachters avoid it like the plague. It seems very strange to me that the area with the highest wind would have the largest temperature increase.
It’s pretty obvious that the climate in most of CA has been cooling and drying since the late 1990s. It’s a general decline. It’s interesting in that it seems to be opposite of much of the Eastern half of the US. The question of phasing must be asked.
Eric Simpson:
They are probably relying on the “Regression to the Mean” falacy. If it is currently cooler than usual, it is almost certain that it will get warmer in the next few years.
If it was 1980 right now and they ran the same model what would the 2012 temperature be? It’s been 32 years so we’re closing in on the 38 year time frame. Oh wait, we must not have added any Co2 to the atmosphere during the past 32 years. That’s the only explanation. /sarc
As I said. 1994 was the turning point from cooling to warming.
mizimi said on June 26, 2012 at 2:06 am:
Last I heard that had been virtually totally outlawed. LA residents would have to become fluorescent with rage.
A few years or so from now, they may have to become LED’d with rage.
This will lead to an unfortunate trend where increasing numbers of evermore enraged LA residents will want the responsible bureaucrats to be “leaded”. I’m hard pressed to not just write that off as “reap what you have sown”.
It’s all about linear extrapolations of sinusoidal variations.
In support of what Bob is saying but slightly further north, the area of Canada that faces the Pacific has shown annual temperature departures linear trend to be dropping from the 1961-1990 averages since 1998. This includes South BC Mountains , Pacific Coast and North BC Mountains and the Yukon. There are fluctations but there is no warming trend here except when El Nino events happen like 2009/2010.
According to NCDC CLIMATE AT A GLANCE web page , the annual temperatures in the US WEST region have been dropping since 1996 and US NORTH WEST region have been dropping since 1992. So the annual temperatures for the entire US western region are dropping for the last 2 decades like in Canada . There are isolated pockets or regions and isolated periods where things are different , but the general trend for entire North American west coast is a slow cooling trend .Bob’s charts for the EASTERN PACIFIC SST ANOMALIES shows a declining temperature trend since 1981, more colder water along the eastern coast of the Pacific than the western part. Is this not what a declining PDO signifies .?
Bob, although a long time ago I wrote numerous hi-end models for potential fields geophysics, later used a number of hydrogeologic models, I just haven’t taken the time to get acquainted with climate models. Are their public versions of the IPCC models that are executable on PCs under Windows 7? If so, a list with links to software and data would be much appreciated. Getting my feet wet and not having to write the code is looking interesting.
The readership please chime in with cogent lists (if such exist).
WC;
Sorry, the models are secret, and numerous, and not for your eyes.