Comments on Riser et al (2016) – An Infomercial about the ARGO Program

UPDATE: I’ve crossed out the entire discussion of the top of the atmosphere energy imbalance and removed the illustration. See the note at the beginning of that heading. 

# # #

An illustration in the paper Riser et al. (2016) Fifteen years of ocean observations with the global Argo array (.pdf copy here.) raised a couple of comments recently at my blog ClimateObservations.  My thanks to bloggers Michael Spurrier and Dave Fair for pointing me to the new paper about the ARGO program.

ABSTRACT

The abstract of Riser et al. (2016) reads like an infomercial:

More than 90% of the heat energy accumulation in the climate system between 1971 and the present has been in the ocean. Thus, the ocean plays a crucial role in determining the climate of the planet. Observing the oceans is problematic even under the most favourable of conditions. Historically, shipboard ocean sampling has left vast expanses, particularly in the Southern Ocean, unobserved for long periods of time. Within the past 15 years, with the advent of the global Argo array of profiling floats, it has become possible to sample the upper 2,000 m of the ocean globally and uniformly in space and time. The primary goal of Argo is to create a systematic global network of profiling floats that can be integrated with other elements of the Global Ocean Observing System. The network provides freely available temperature and salinity data from the upper 2,000 m of the ocean with global coverage. The data are available within 24 hours of collection for use in a broad range of applications that focus on examining climate-relevant variability on seasonal to decadal timescales, multidecadal climate change, improved initialization of coupled ocean–atmosphere climate models and constraining ocean analysis and forecasting systems.

FIGURE 5 FROM RISER ET AL.

I’ve included Figure 5 from Riser et al. (2016) as my Figure 1.

Figure 1

 

The caption cites two references. Reference 27 is Roemmich et al. (2015) Unabated planetary warming and its anatomy since 2006, which is an ARGO array-based paper. And reference 42 is Levitus et al. (2012) World ocean heat content and thermosteric sea level change (0–2,000 m), 1955–2010., which is the paper that supports the ocean heat content data from the NOAA National Oceanographic Data Center (NODC).  Basically, for their Figure 5, Riser et al. overlaid the (black, red and blue) curves of the ARGO-based ocean heat content data (0-2000 meters) for 2006-2014 atop the corresponding long-term pentadal data from the NODC (purple curve), and then claimed victory because the two datasets showed similar warming rates.

Riser et al. (2016) states (my boldface):

In Fig. 5 the Argo-only plots (inset) are only for the years 2006–2014, the period when global coverage from Argo exists. So these short plots are overlaid on a plot of heat content estimates for 0–2,000 m depth for the period 1955–2010₄₂. The Argo estimates show a very similar trend. This is a crucial result in making an assessment of the ocean’s role in climate change, one that would have probably been impossible before Argo.

Oddly, Riser et al. (2016) do not list the trend values and their uncertainties for the ARGO-based data so we can’t compare them to the NODC data over any timeframe. I’ll let you speculate about that.  But for the period where they overlap, the ARGO data and the NODC data should show similar trends. The NODC uses ARGO-based data for their ocean heat content estimates.

Let’s zoom in on the inset graph of the ARGO-based data included in Figure 5 from Riser et al. (2016), my Figure 2.

Figure 2

Obviously, based on the ARGO floats, the vast majority of the ocean heat uptake for the depths of 0-2000 meters has taken place in the extratropics of the Southern Hemisphere (60S-20S/black curve) for the period of 2006-2014.  Riser et al. (2016) acknowledge this in the text (my boldface):

The data have also allowed temporal spa­tial variations in ocean heat content to be discerned40, suggest­ing that most of the increase in heat content in the past decade has occurred in the Southern Ocean (which was poorly sampled before Argo); it has also been noted41 that ENSO variability in the tropical Pacific has for now somewhat obscured the global increase in sea surface temperature.

The NOAA/NODC ocean heat content for the same depths and timeframe also shows the additional ocean heat uptake in the extratropics of the Southern Hemisphere. See my Figure 3. The NODC ocean heat content data being presented are available from the KNMI Climate Explorer.

Figure 3

You’ll note in Figure 2 that Riser et al have presented the data in units of Joules*10^22, while for Figure 3 I’ve used the NODC data at the KNMI Climate Explorer, which are presented in Gigajoules per square meter (GJ/m^2). That’s of no concern here, because Riser et al. did not present the trend values and their uncertainties.  So let’s simply look at whether the trend lines are positive or negative.

Curiously, for the same period and depths, the NODC data (Figure 3) are showing warming of the extratropics of the Northern Hemisphere comparable to the tropics, while the ARGO data (Figure 2) show a slight cooling in the Northern Hemisphere. With NOAA’s history of manipulating ocean surface temperature data to create warming, many readers will not find that surprising.

CLIMATE MODEL-BASED TOA ENERGY IMBALANCE CANNOT EXPLAIN HEAT UPTAKE IN THE EXTRATROPICS OF THE SOUTHERN HEMISPHERE

UPDATE: In my haste to provide a graph of top of the atmosphere energy imbalance that was easy to visually compare to ocean heat uptake, I converted the energy imbalances for the three regions to anomalies. That was a mistake. The modeled tropical energy imbalance at the top of the atmosphere is positive, while the modeled energy imbalances in the extratropics of both hemispheres are negative values. So I’ve crossed out the remainder of the discussion of energy imbalances and deleted that graph. Thanks to Nicholas Schroeder for the reminder here.

The climate model-based ocean heat uptake is often calculated by the climate science community from the modeled energy imbalance at the top of the atmosphere.  See Chapter 1.24 – A Rough Calculation of the Amount of Missing Heat…A Critical Issue from my free ebook On Global Warming and the Illusion of Control – Part 1 (25MB .pdf).

Figure 4 presents the climate model-simulated energy imbalance at top of the atmosphere (TOA), based on the CMIP5 climate models with the RCP6.0 forcings, for the period of 2006-2014. Climate models show the greatest increase in the TOA energy imbalance in the tropics.  And you’ll note the models are also showing a higher increase in extratropics of the Northern Hemisphere than the extratropics of the Southern Hemisphere.  One might conclude from the disparities in relative trends between Figures 3 and 4 that naturally occurring ocean processes are responsible for the additional ocean heat uptake of the extratropics of the Southern Hemisphere.

Figure 4

The top of the atmosphere energy imbalance is calculated by subtracting the modeled Outgoing Shortwave Radiation (rsut) and Outgoing Longwave Radiation (rlut) from Incident Shortwave Radiation (rsdt).   The acronyms are the nomenclature found at the Monthly CMIP5 scenario runs webpage at the KNMI Climate Explorer.  I’ve presented the top of the atmosphere energy imbalance in anomaly form to highlight the trends.

We already know that climate models cannot simulate the spatial patterns of the warming of the surfaces of the global oceans.  See the recent post Climate Models Are NOT Simulating Earth’s Climate – Part 1. If the models cannot explain where and when the oceans warm at the surface, is it safe to assume the models do not properly simulate the warming of the oceans to depth?

CLOSING

I’ve only addressed a small portion of Riser et al. (2016) in this post.  Feel free to comment on additional portions that I’ve overlooked.

Riser et al. (2016) claim the agreement between the ARGO-based ocean heat content data and the NODC data is important.  But they failed to present those trends and their uncertainties. Regardless, the NODC uses ARGO-based data so the NODC and ARGO-based data should agree where they overlap. Not too surprising:  the NODC show a noticeable ocean heat uptake for the extratropics of the Northern Hemisphere (20N-60N) for the period of 2006-2014, while the ARGO data show a slight heat loss.

The climate model-simulated energy imbalance at the top of the atmosphere cannot explain the extra ocean heat uptake in the extratropics of the Southern Hemisphere as shown by the ARGO-based data. The simulated energy imbalance for the tropics shows the highest increase while the imbalance for the extratropics of the Southern Hemisphere is lowest.  Thus, the additional ocean heat uptake in the extratropics of the Southern Hemisphere must be due to ocean processes.

Sadly, climate models fail to properly simulate the spatial patterns of warming of ocean surfaces.  Can the climate models used by the IPCC for their 5^th Assessment Report explain the extra ocean heat uptake in the extratropics of the Southern Hemisphere as shown by the ARGO-based data for the period of 2006-2014 or explain the relative lack of ocean heat uptake in the tropics and extratropics of the Northern Hemisphere during the ARGO era?  Unfortunately, the outputs of the climate model-simulated ocean heat uptake for the CMIP3- and CMIP5-archived models are not available in easy-to-use form.