A bit of a tiff developed over at Dr. Roger Pielke’s place over disagreements on the recent Texas heatwave being attributed to AGW or to ENSO. Bob Tisdale has something to say about that. Bob writes:
“In one email, Roger referred to my post about how poorly the new NCAR model hindcasts certain temperature indices, including ENSO, and Nelsen-Gammon’s decided to call my discussion about ENSO a red herring. Little does he know, I have observation-based data to back my claims.”
by Bob Tisdale
I can see no basis for John Nielsen-Gammon’s attempt to attribute the record high temperatures in Texas to the hypothesis of Anthropogenic Global Warming. It appears that Nielsen-Gammon, like the Intergovernmental Panel on Climate Change (IPCC), relies on climate models to conclude that most of the rise in Surface Temperatures, globally and regionally, is caused by anthropogenic greenhouse gases. Unfortunately, their reliance on models to support that hypothesis is unfounded. The climate models show little to no skill at hindcasting past global and regional natural variations in Sea Surface Temperature, which, through coupled ocean-atmospheric processes, would have impacts on the temperature and drought in Texas. Since the climate models are incapable of replicating the natural modes of multiyear and multidecadal variability in Sea Surface Temperatures, the models are of little value as tools to determine if the warming could be attributed to manmade or natural causes, and they are of little value as tools to project future climate on global or regional bases.
And based on John Nielsen-Gammon’s comment about El Niño-Southern Oscillation (ENSO), it appears he has overlooked the significant contribution ENSO can make to the multiyear and multidecadal variations in Global Sea Surface Temperature anomalies, which are so obvious during the satellite-era of Sea Surface Temperature observations.
Roger Pielke Sr., has published at his blog a series of emails between he and John Nielsen-Gammon. Roger’s post is dated November 10, 2011 and is titled John Nielsen-Gammon and I Continue Our Discission. Pielke Sr.’s initial post on this topic, dated November 4, 2011, is titled NBC Nightly News Regarding The Recent October Snowstorm And A Quote From John Nielsen-Gammon. In it, Pielke Sr. refers to Nielsen-Gammon’s September 9, 2011 blog post at the Houston Chronicle website Chron.com titled Texas Drought and Global Warming. All three posts are worth a read and provide the fuel for this post.
In one of the emails reproduced in his recent post, Roger Pielke Sr. provided Nielsen-Gammon with a link to my November 4, 2011 post An Initial Look At The Hindcasts Of The NCAR CCSM4 Coupled Climate Model. (Please read this post also, if you haven’t done so already. It shows how poorly the recent version of the NCAR CCSM coupled climate model replicates the surface temperatures from 1900 to 2005.) And Nielsen-Gammon’s response to it included:
“When driven by observed oceanic variability, the models do a great job simulating the atmospheric response. With the present drought, it’s not a matter of predicting the oceans and atmosphere. We know the present ocean temperature patterns, so we can estimate their contribution very well from both observations and models. The models’ difficulty in simulating the statistics of ENSO itself is a red herring.”
First, I have no basis from which to dispute Nielsen-Gammon’s opening sentence of, “When driven by observed oceanic variability, the models do a great job simulating the atmospheric response”. I have not investigated how well the models actually perform this function. But that’s neither here nor there. Why? Well, if the hindcast and projected representations of sea surface temperatures created by the models are not realistic, then the atmospheric response to the modeled oceanic variability would also fail to be realistic.
Second, Nielsen-Gammon wrote, “We know the present ocean temperature patterns, so we can estimate their contribution very well from both observations and models.” Nielsen-Gammon’s sentence does not state that the models provide a reasonable representation of ocean variability. So the fact that Nielsen-Gammon can estimate the oceanic contributions from observations AND from models is immaterial. The models are so far from reality, they have little value as climate hindcasting, or projection, or attribution tools, as stated previously.
Also, if you’re new to the subject of climate change, always keep in mind, when you read a climate change post like John Nielsen-Gammon’s, where the author constantly refers to models and model-based studies (in an attempt to add credibility to the post?), that it may not be the same climate model being referred to. Models have strengths and weaknesses, and climate scientists use different models for different studies. Depending on the coupled ocean-atmosphere process being studied, even if one organization’s model is used, model parameters may be set differently, they may be initialized differently, they may use different forcings, etc. So, while two model-based climate studies may use the same model, the model runs used to study the atmospheric response to the Atlantic Multidecadal Oscillation, for example, may not incorporate the same forcings that are used to hindcast past climate and project future climate. In fact, there are model-based studies where observed Sea Surface Temperature data are used to force the climate models.
MORE EXAMPLES OF HOW POORLY CLIMATE MODELS DEPICT SEA SURFACE TEMPERATURE VARIATIONS
In addition to the post linked earlier in which I compared climate model outputs to observed data, I have also illustrated and discussed in detail the differences between the observed sea surface temperature anomalies and those hindcast/projected by climate models in the two posts titled:
In those posts, I showed the very obvious differences between observed Sea Surface Temperature data and the model mean of the climate models used in the IPCC AR4 on global and ocean-basin bases, during the satellite-era of sea surface temperature measurement, 1982 to present. Here are a few examples:
Figure 1 is a time-series graph of the satellite-based observations of Global Sea Surface Temperatures versus the model mean of the hindcasts/projections made by the climate models used in the IPCC AR4. It shows how poorly the linear trend of the model mean compares to the trend for the measured Global Sea Surface Temperature anomalies. The models overestimate the warming by approximately 50%.
Figure 2 compares the linear trends for the observations and the model mean of the IPCC AR4 hindcasts/projections of Sea Surface Temperatures on a zonal mean basis. That is, it compares, for the period of January 1982 to February 2011, the modeled and observed linear trends, in 5-degree-latitude bands (80S-75S, then 75S-70S, etc., from pole to pole) from the Southern Ocean around Antarctica north through to the Arctic Ocean. It clearly shows that, in the models, the tropics warm faster than at higher latitudes, where in reality, that is clearly not the case. This implies that the models do an extremely poor job of simulating how the oceans distribute warm water from the tropics toward the poles. Extremely poor.
In those two posts, I not only illustrate the failings of the models on a Global basis, but I also illustrate them on an ocean-basin basis: North and South Pacific, East and West Pacific, North and South Atlantic and Indian Ocean. There are no subsets of the models that come close to the observations on a time-series basis and on a zonal-mean basis.
John Nielsen-Gammon notes in his article, after he changed attribution from “greenhouse gases” to “global warming”, that:
The IPCC has not estimated the total century-scale contribution to global warming from anthropogenic greenhouse gases, but has said that the warming since 1950 was probably mostly anthropogenic. So it seems reasonable to estimate that somewhere around two-thirds of the century-scale trend is due to anthropogenic greenhouse gas increases. That is to say, the summer temperatures would have been about
one or one and a half degrees coolerone half to one degree cooler without the increases in CO2 and other greenhouse gases. [John Nielsen-Gammon’s boldface and strikes.]
I cannot see how Nielsen-Gammon can make that claim when the IPCC’s model depictions of sea surface temperature variability over the past 30 years, which are coupled to global and regional variations in temperature and precipitation, differ so greatly from the observations. I truly cannot. The models are so different from observations that they have no value as an attribution tool. None whatsoever.
ON ENSO BEING A RED HERRING
The last sentence in the first quote from John Nielsen-Gammon above reads, “The models’ difficulty in simulating the statistics of ENSO itself is a red herring.” As a reference, Animation 1, is the El Niño-Southern Oscillation (ENSO)-related comparison from my post that Roger Pielke Sr. linked for Nielsen-Gammon (An Initial Look At The Hindcasts Of The NCAR CCSM4 Coupled Climate Model).It shows how poorly the models hindcast the frequency, magnitude, and trend of ENSO events. In that post, I explained why the failure of climate models to reproduce the frequency and magnitude of ENSO events was important. Yet John Nielsen-Gammon characterized my illustrations and discussion as a “red herring”.
Here’s what I wrote, in part, about Animation 1:
The first thing that’s obviously different is that the frequency and magnitude of El Niño and La Niña events of the individual ensemble members do not come close to matching those observed in the instrument temperature record. Should they? Yes. During a given time period, it is the frequency and magnitude of ENSO events that determines how often and how much heat is released by the tropical Pacific into the atmosphere during El Niño events, how much Downward Shortwave Radiation (visible sunlight) is made available to warm “and recharge” the tropical Pacific during La Niña events, and how much heat is transported poleward in the atmosphere and oceans, some of it for secondary release from the oceans during some La Niña events. If the models do not provide a reasonable facsimile of the strength and frequency of El Niño and La Niña events during given epochs, the modelers have no means of reproducing the true causes of the multiyear/multidecade rises and falls of the surface temperature anomalies. The frequency and magnitude of El Niño and La Niña events contribute to the long-term rises and falls in global surface temperature.
My illustrations and discussions of ENSO in that post are not intended to divert anyone’s attention from the actual cause of the rise in global temperatures, which is what I assume John Nielsen-Gammon intended with the “red herring” remark. The frequency and magnitude of ENSO events are the very obvious cause of the rise in Sea Surface Temperatures during the satellite era. And that isn’t a far-fetched hypothesis; that is precisely the tale told by the sea surface temperature data itself. One simply has to divide the data into logical subsets to illustrate it, and it is so obvious once you know it exists that it is hard to believe that it continues to be overlooked by some members of the climate science community.
Recently I started including two illustrations of ENSO’s effect on Sea Surface Temperatures in each of my monthly Sea Surface Temperature anomaly updates. (Example post: October 2011 Sea Surface Temperature (SST) Anomaly Update) Refer to the graphs of the “volcano-adjusted” East Pacific Sea Surface Temperature anomalies and of the Sea Surface Temperature anomalies for the Rest of the World. I’ve reposted them here as Figures 3 and 4, respectively.
Note Regarding Volcano Adjustment: I described the method used to determine the volcano adjustment in the post Sea Surface Temperature Anomalies – East Pacific Versus The Rest Of The World, where I first illustrated these two datasets. The description reads:
To determine the scaling factor for the volcanic aerosol proxy, I used a linear regression software tool (Analyse-it for Excel) with global SST anomalies as the dependent variable and GISS Stratospheric Aerosol Optical Thickness data (ASCII data) as the independent variable. The scaling factor determined was 1.431. This equals a global SST anomaly impact of approximately 0.2 deg C for the 1991 Mount Pinatubo eruption.
Back to the discussion of the volcano-adjusted East Pacific and Rest-of-the-World data: Let’s discuss the East Pacific data first. As you’ll quickly note in Figure 3, based on the linear trend produced by EXCEL, there has been no rise in the Sea Surface anomalies for the volcano-adjusted East Pacific Ocean Sea Surface Temperature anomaly data, pole to pole, or the coordinates of 90S-90N, 180-80W, for about the past 30 years. The El Niño events and La Niña events dominate the year-to-year variations, as one would expect, but the overall trend is slightly negative. The East Pacific Ocean dataset represents about 33% of the surface area of the global oceans, and there hasn’t been a rise in sea surface temperature anomalies there for three decades.
Since we’ve already established that Global Sea Surface Temperature observations have risen during that period (Refer back to the observation-based data in Figure 1), that means the Rest-of-the-World data is responsible for the rise in global Sea Surface Temperature anomalies. But as you’ll note in Figure 4, the volcano-adjusted Sea Surface Temperature anomalies for the Rest of the World (90S-90N, 80W-180) rise in very clear steps, and that those rises are in response to the significant 1986/87/88 and 1997/98 El Niño/La Niña events. (It also appears as though the Sea Surface Temperature anomalies of this dataset are making another upward shift in response to the 2009/10 El Niño and 2010/11 La Niña.) And between those steps, the Rest-of-the World Sea Surface Temperature anomalies remain relatively flat. How flat will be illustrated shortly.
Note: The periods used for the average Rest-Of-The-World Sea Surface Temperature anomalies between the significant El Niño events of 1982/83, 1986/87/88, 1997/98, and 2009/10 are determined as follows. Using the NOAA Oceanic Nino Index(ONI) for the official months of those El Niño events, I shifted (lagged) those El Niño periods by six months to accommodate the lag between NINO3.4 SST anomalies and the response of the Rest-Of-The-World Sea Surface Temperature anomalies, then deleted the Rest-Of-The-World data that corresponds to those significant El Niño events. I then averaged the Rest-Of-The-World SST anomalies between those El Niño-related gaps.
I have in numerous posts discussed, illustrated, and animated the variables associated with the coupled ocean-atmosphere process of El Niño-Southern Oscillation (ENSO) that cause these apparent upward shifts in the Rest-of-the-World Sea Surface Temperature anomalies. My first posts on this were in January 2009. The most recent ones are from the July 2011: ENSO Indices Do Not Represent The Process Of ENSO Or Its Impact On Global Temperature and Supplement To “ENSO Indices Do Not Represent The Process Of ENSO Or Its Impact On Global Temperature”.Those two posts were written at an introductory level for those who aren’t familiar with the process of the El Niño-Southern Oscillation (ENSO). In the initial post, I further illustrated the actual linear trends of the Rest-of-the-World data between the significant ENSO events, reproduced here as Figure 5. They are indeed flat.
And in the supplemental post, I further subdivided the Rest-of-the-World Sea Surface Temperature data into two more subsets. The first to be illustrated, Figure 6, covers the South Atlantic, Indian and West Pacific Oceans. As shown, Sea Surface Temperature anomalies decay between the significant ENSO events, just as one would expect.
And for the North Atlantic, Figure 7, which is impacted by another mode of natural variability called the Atlantic Multidecadal Oscillation (AMO), the linear trends between those significant ENSO events are slightly positive, also as one would expect. And the short-term ENSO-induced upward shifts are plainly visible in Figure 7 and are responsible for a significant portion of the rise in North Atlantic Sea Surface Temperature anomalies over the past 30 years.
This post clearly illustrates that John Nielsen-Gammon failed to consider that climate models prepared for the Intergovernmental Panel on Climate Change (IPCC) AR4 have little to no basis in reality. When one considers the significant differences between the observed Sea Surface Temperature anomaly variations and those hindcast/projected by climate models, the models provide no support for his conclusion that most of the rise in Surface Temperatures, globally and regionally, was caused by anthropogenic greenhouse gases.
This post also clearly illustrated that “The models’ difficulty in simulating the statistics of ENSO itself is”…NOT…“a red herring.” The process of the El Niño-Southern Oscillation was responsible for most of the rise in global sea surface temperature anomalies over the past thirty years.
For the sources of data presented in this post, refer to the linked posts from which the graphs were borrowed.