Ocean Surface Temperature Limit-Part 1

By Richard Willoughby

May 2021

(The author appreciates the availability NASA’s Earth Observations satellite data sets used in this analysis.)

This is a three part series that analyses the role of atmospheric water in regulating Earth’s thermal balance.

Part 1 is an analysis of the temperature of tropical ocean warm pools and the temperature limiting processes.

Part 2 discusses the mechanism of deep convection concluding with the persistency of clouds over ocean warm pools.

Part 3 examines the global ocean energy balance over an annual cycle month-by-month to identify the role of atmospheric water in regulating the energy balance.

Part 1:  Observed Ocean Surface Temperature Regulation – Upper Limit

Ocean surface temperature observations show that less than 10% of open ocean surface water exceeds an annual average surface temperature of 30C.  Also less than 1% of the ocean surface exceeds 32C for more than a few days.  The thermostatic upper limit of 30C for ocean warm pools has been observed scientifically since at least the 1970s and is easily observed globally in the current era using satellite measurements.  Each year, warm pools reach their maximum extent of around 9% of the ocean water surface area in April as both Tropical Western Pacific and Indian Ocean reach peak heat uptake.  This is observed with reference to Figure 1.

Figure 1: Global map highlighting tropical warm pools at or above30C shown in red – April 2020.

The Northern Hemisphere continues to warm from April through to August.  The Pacific warm pools contract as they advance northward and the Indian Ocean warm pools dissipate as the northern monsoon sets in over the Arabian Sea and Bay of Bengal while the Gulf of Mexico becomes a temperature regulating warm pool and the Persian Gulf becomes the warmest ocean surface on the globe.  Figure 2 highlights this transition by comparison with Figure 1.

Figure 2: Extent of warm pools in August 2019 – red above 30C, yellow above 34C

By September, the solar zenith has moved south and the global oceans are in a cooling phase.  The Gulf of Mexico and Persian Gulf cool rapidly.  Ocean warm pools reduce to their minimum extent in September with only 3% of the ocean water surface at or above 30C as depicted in Figure 3.

Figure 3: Warm pool minimum extent in September – September 2019

The surface temperature upper limit of 32C becomes more apparent when the proportion of ocean surface area at any given temperature is charted as shown in Figure 4.

Figure 4:  Area of ocean surface at a specific temperature as a proportion of the total ocean surface area and cumulative proportion by area at or below given surface temperature.

More than 50% of the ocean water surface exceeds 22C as shown in Figure 4 right panel while the peak proportion of area is in the range 28C to 29C but falls off sharply above 29C such that there is less than 1% of the ocean surface area warmer than 32C, left panel.

Warm Pool Temperature Limiting Process Control

The response of the temperature limiting process over tropical warm pools can be observed in hourly intervals when the warm pools exist at the tropical moored buoys.  The same temperature limiting process is observed across the three tropical oceans as set out in the series of ocean surface temperature limiting charts in Figure 5.

Figure 5:  Temperature limiting process in the tropical warm pools; top Pacific, middle Atlantic and bottom Indian (Charted data sourced from NOAA Pacific Marine Environment Laboratory).

With reference to the surface temperature control charts in Figure 5, the temperature limiting process begins when the surface temperature rises above 30C in clear sky conditions before the local convective cycle sets in.  The cycle starts with cloudburst followed by persistent cloud that reduces surface insolation to slow the rate of rise in temperature.  Moist air diverges from cooler adjacent zones to the warm pool resulting in heavy precipitation that increases the cooling rate in the warm pool.  The high level moist air has been cooled over water at 28C to 29C that is absorbing heat at the surface due to predominantly clear sky conditions above.  Convergence of moist air to the warm pool continues until it is no longer the warm pool.  This is particularly apparent in the middle image where the Atlantic warm pool cools below 30C from day 22 and the precipitation drops to zero as the warm pool moves on.

The Annual Cycle of Warm Pools

As previously observed the warmest regions of the tropical oceans can move appreciably throughout the year but the Equator usually observes the highest annual average surface temperature; always close to 30C in all warm pools as observed in the series of south to north transects through warm pools displayed in Figure 6.

Figure 6:  Monthly and annual average sea surface temperature south to north transects – top Western Pacific, middle Indian Ocean through the Arabian Sea, bottom Gulf of Mexico and Eastern Pacific

The Persian Gulf becomes the warmest ocean surface in August each year.  It is the only ocean surface that regularly exceeds 34C in any year.  This unique deviation from the temperature regulating feature of the tropical oceans is highlighted in Figure 7.

Figure 7:  Monthly and annual average sea surface temperature south to north transects centred on 53.5W through the Persian Gulf and Indian Ocean – August highlighted with red curve.

Performance of Climate Models in predicting Ocean Surface Temperature

The tropical warm pools present a challenge for climate models because the models rely on cloud parameterisation rather than the actual physics of convective instability that limits the maximum ocean surface temperature.  The US National Oceanic and Atmospheric Administration provides a high quality data set for sea surface temperature in the tropical Pacific that combines ocean water temperature measurements observed at the tropical moored buoys interpolated between buoys using satellite imagery.  Figure 8 demonstrates the poor performance of the CISIRO and Bureau of Meteorology model, Australian Community Climate and Earth Systems (ACCESS) climate model over the Nino34 region in the tropical Pacific.  The Nino34 region is located between latitudes 5N and 5S from longitudes 120W to 170W.  This region is recognised as an important indicator of global weather and particularly the Pacific El Nino and La Nina modes that influence weather conditions across the eastern States of Australia. 

Figure 8:  Comparison of ACCESS climate model prediction and hindcast over five decades with the actual observed ocean surface temperature over four decades.

It is quite apparent that the model has cooled the past to an extreme degree to maintain its warming trend.  According to the ACCESS hindcast, the well-known 1998 El Nino in Australia could not have possibly occurred because the temperature in this region is observed to exceed 28C for El Nino conditions to form as shown in Figure 8 in the measured data.  The model output clearly demonstrates its poor performance over this important weather influencing region.

The uniqueness of the Persian Gulf provides another example of the poor performance of the ACCESS model over this region in a comparison with measured data in Figure 9.

Figure 9:  Comparing measured and modelled surface temperature in the Persian Gulf for the month of August 1960 to 2020.

In August each year, the Persian Gulf becomes the warmest ocean surface on the globe.  The unique conditions of low mid-level moisture prevents the formation of convective instability and the surface temperature regulating phenomena associated with high level convective cloud.  Here the ACCESS model again produces a poor result but in this case up to four degrees Centigrade cooler than the measured data.

Data Sources

https://neo.sci.gsfc.nasa.gov/view.php?datasetId=CERES_LWFLUX_Mhttps://neo.sci.gsfc.nasa.gov/view.php?datasetId=CERES_SWFLUX_Mhttps://neo.sci.gsfc.nasa.gov/view.php?datasetId=MYD28Mhttps://neo.sci.gsfc.nasa.gov/view.php?datasetId=MYD28M

I have used many months from these sets.  All the charts and images are independently produced meaning not copied images from these links.

There is also data from the moored buoys that I refer to:https://www.pmel.noaa.gov/tao/drupal/disdel/