By Andy May
In previous posts, see here and here, I’ve tried to show that because the oceans cover 71% of Earth and they contain 99% of the thermal energy stored on the Earth’s surface, they dominate the speed and magnitude of climate changes. In all my posts the Earth’s surface is defined as everything from the ocean floor to the top of the atmosphere. The details of the calculation of ocean and atmospheric heat content is detailed in this spreadsheet. The ocean’s huge heat capacity prevents large temperature swings and dampens and delays those that do occur.
Attempting to show the direction, speed, and magnitude of climate change by measuring and averaging atmospheric surface temperatures is pointless, in my opinion. The record we have of atmospheric and ocean surface temperatures is too short and far too inaccurate to provide us with useful trends on a climatic (30 years +) time scale. Further, these records are sporadic measurements in a chaotic surface zone that has large temperature swings. In Montana, United States, for example, recent minimum/maximum temperatures have been as low as -70°F (-57°C) and as high as 117°F (47°C). These enormous swings make measuring year-to-year global average differences of 0.1°C exceedingly difficult. Yet, this is the precision demanded if we are to properly characterize a climate that is only warming at a rate of roughly 1.4°C/century, which is 0.014°C per year and 0.14°C/decade.
The measurements are especially useful for predicting the weather but are inappropriate for measuring changes of less than half of a degree over climatic time spans. We need to measure something more stable and less chaotic for this purpose. This post and the next one present evidence that the mixed layer in the ocean seems to be well suited to the task.
In my opinion, to get a proper handle on the direction and speed of global warming, we must look to temperature changes in the ocean. Especially those portions of the ocean that are in constant contact with the atmosphere. Very long-time frame climate change, one-thousand years plus, involves the entire ocean. But for time frames of one-hundred years or less, we are dealing mostly with the upper few hundred meters of the ocean.
The temperature profile of the upper ocean is very complex. This is complicated by the poor quality of our ocean-surface temperature measurements, especially prior to the introduction of Argo floats and modern ocean buoy measurements, like the Triton buoys over the last 20 years. Ships cover a limited area of the ocean and the depth, consistency and quality of their temperature measurements are uncertain. Satellite measurements of the very top of the ocean are possible, but these measurements are complicated by what is called the ocean skin effect.
The Ocean Skin
At the ocean-air interface, temperatures change rapidly. The magnitude of the change and the thickness of the uppermost ocean affected is determined by cloudiness, whether it is night or day, and wind speed. This “skin” is thicker on calm cloudless days and thinner at night and on windy cloudy days. The temperature at the ocean-air interface (“SST”) is what is measured by radiometers and satellites. Unfortunately, the relationship between this temperature and the more stable mixed-layer temperature or “foundation” temperature is unknown. The relationship changes rapidly and is complicated. Various models have been proposed (Horrocks, O’Carroll, Candy, Nightingale, & Harris, 2003), but none have the reliability and accuracy required.
To make matters worse, right at the surface there is a population of cyanobacteria that works to change the temperature and lower the salinity of the surface water (Wurl, et al., 2018). The sea surface temperature problem is best illustrated with the diagram in Figure 1, from GHRSST or the Group for High Resolution Sea Surface Temperature. They are striving to understand the ocean skin layer so that satellite sea-surface temperature measurements can be properly combined with measured ocean temperatures.
The temperature difference between the SST and the stable portion of the mixed layer can be three to six degrees daily (Wick & Castro, 2020). As Gary Wick and Sandra Castro explain:
“The daily cycle of solar radiation leads to periodic heating of the near-surface layer of the ocean. At low wind speeds, turbulent mixing is reduced, and a warm layer and diurnal thermocline can form near the ocean surface during the day. At night, mixing typically erodes this layer. While the amplitude of diurnal warming is relatively small on average [0.5 K], under conditions of very low wind speed and sufficient insolation, the warming at the surface sensed by satellites can be highly significant… In situ observations have shown warming in excess of 5K at depths of 0.3–0.6 m. Satellite observations from multiple sensors have observed extreme warming events up to 7K in magnitude at the surface, and it has been suggested that events exceeding 5K are not infrequent.” (Wick & Castro, 2020)
The temperatures in the quote are given in Kelvin (K), and equivalent to degrees Celsius. The main point is that exceptionally large differences in the ocean SST occur in calm conditions on clear (cloudless) days. Figure 1 shows that the temperature increases can affect water as deep as ten meters. But differences of more than 0.5°C are almost always limited to the top meter of the ocean. As we will see in the next post on the mixed layer, these known skin anomalies are ignored in ocean temperature datasets. They often have a measurement labeled as zero depth, but it is taken under the surface, usually at a depth of 20 cm or more. The mixed-layer temperature is often defined as the temperature of the layer that has a temperature within 0.5°C of the surface temperature (Levitus, 1982). This is not precise, what they mean is the temperature of the ocean just under the surface, perhaps 20 to 100 cm. Except on clear windless days, this will be the “foundation” temperature. At night and on cloudy or windy days, the temperature will always be the “foundation” temperature.
The mixed layer has homogenous properties due to turbulent mixing and using a temperature difference limit of 0.5°C is a convenient definition, but it breaks down near the poles in the winter where a more complex definition is needed. Numerous methods have been proposed, too many to list here, but the complex one described by James Holte and Lynne Talley (Holte & Talley, 2008) is currently favored. Their technique is widely used today to choose a “mixed-layer depth,” which is the bottom of the mixed layer. It is necessary because in the polar regions, in the winter, deep convection, driven by surface heat loss, can mix the water column to 2000 meters or even deeper. As we will see in the next post, it is in these areas that thermal energy from the surface is transmitted to the deep ocean.
There are many ocean temperature datasets, and we will discuss the data from several of them in the next post. Figure 2 is a plot of the global average December ocean temperatures from the surface to 140 meters from the University of Hamburg datasets. This plot illustrates the temperature profile terms we have discussed, with real global data.
The temperatures reported by the University of Hamburg are average temperatures over more than 12 years and do not represent any given year. The NOAA MIMOC temperatures, which we will look at in the next post, are the same. Figure 3 shows the average year of the measurements and the standard deviation of the years.
Both the University of Hamburg and NOAA recognize that the Argo data, which is the bulk of their raw data, is sparse. There is one float per 3° of latitude and longitude (~32,913 sq. miles at 40° North or South or 84,916 sq. km.) This float sends one complete profile to us every ten days. The University and NOAA have decided that to combat the paucity of data, they should compile monthly averages of all data they could find. As we will see in the next post the big changes in the mixed layer occur by month and location, so this makes some sense.
Above the foundation or mixed layer, there are other zones identified in Figure 1. These are defined by GHRSST as follows. I’ve edited the GHRSST text for clarity, the original text can be viewed here.
The interface temperature (SSTint)
At the exact air-sea interface a hypothetical temperature called the interface temperature (SSTint) is defined although this is of no practical use because it cannot be measured using current technology.
The skin sea surface temperature (SSTskin)
The skin temperature (SSTskin) is defined as the temperature measured by an infrared radiometer typically operating at wavelengths 3.7-12 µm (chosen for consistency with the majority of infrared satellite measurements) that represents the temperature within the conductive diffusion-dominated sub-layer at a depth of ~10-20 micrometers. SSTskin measurements are subject to a large potential diurnal cycle including cool skin layer effects (especially at night under clear skies and low wind speed conditions) and warm layer effects in the daytime.
The sub-skin sea surface temperature (SSTsub-skin)
The subskin temperature (SSTsubskin) represents the temperature at the base of the conductive laminar sub-layer of the ocean surface. For practical purposes, SSTsubskin can be well approximated to the measurement of surface temperature by a microwave radiometer operating in the 6-11 GHz frequency range, but the relationship is neither direct nor invariant to changing physical conditions or to the specific geometry of the microwave measurements.
The surface temperature at depth (SSTz or SSTdepth)
All measurements of water temperature beneath the SSTsubskin are referred to as depth temperatures (SSTdepth) measured using a wide variety of platforms and sensors such as drifting buoys, vertical profiling floats (like Argo), or deep thermistor chains at depths ranging from 10 to 750m (like Triton and Tao). These temperature observations are distinct from those obtained using remote sensing techniques (SSTskin and SSTsubskin) and must be qualified by a measurement depth in meters.
The foundation temperature (SSTfnd)
The foundation SST, SSTfnd, is the temperature free of diurnal (daily) temperature variability. That is, the top of the stable portion of the mixed layer. Only in situ contact thermometry can measure SSTfnd.
In summary, SST and atmospheric surface temperatures are too affected by weather and diurnal variability to measure climate changes with any reliability or precision. Total ocean or deep ocean temperatures give us a hint at long climatic changes of a thousand years or more, but they tell us little about changes in the one-hundred-year range.
The ocean mixed layer is a zone that begins between one millimeter and roughly ten meters below the ocean surface. Above this depth, temperatures are affected minute-by minute by the atmosphere and sunlight. At night, the top of the mixed layer moves closer to the surface but can be affected by windspeed, precipitation, and cloudiness. Below the top of the mixed layer, the temperature is more stable than the atmosphere and ocean surface. The temperature, salinity, and density of the layer is nearly constant from the top to the bottom due to turbulence. It reflects surface temperatures but is a function of an average of the previous several weeks. The thickness of the mixed layer varies seasonally from a few tens of meters to several hundred meters. We will discuss the mixed layer in much more detail in the next post.
None of this is in my new book Politics and Climate Change: A History but buy it anyway.
You can download the bibliography here.