Issue 041 - Climate - Ocean heat content
How much energy does an unusually warm Mediterranean contain?
Reuters reported that July 2026 brought record heat to European seas, while a World Weather Attribution analysis estimated that human-caused climate change has added about 2 degrees C of warming to Mediterranean sea-surface temperatures.
The problem
Estimate how much additional thermal energy is represented by warming the Mediterranean's upper ocean by 2 degrees C.
Choose a reasonable surface-layer depth rather than assuming the entire Mediterranean has warmed uniformly to great depth. Then compare the stored heat with a familiar energy scale such as annual global electricity generation, total human energy consumption, or the output of a large power plant over time.
Because Fermi problems target an order of magnitude, I normally use no more than two significant digits and write most calculations in scientific notation; the Fermi reference explains both conventions.
Before checking sources
Matt's first pass
I assumed the Mediterranean Sea has a surface area about twice the surface area of Texas. I used Texas as about 10^12 m2, so the Mediterranean would be about:
Mediterranean area ~= 2 x 10^12 m2I also used my existing peg that humans use about 5.8 x 10^20 J per year worldwide.
Starting with only the top meter of water, the volume is:
volume ~= area x depth
~= 2 x 10^12 m2 x 1 m
~= 2 x 10^12 m3
~= 2 x 10^15 LIf it takes about 4 x 10^3 J to increase a liter of water by 1 degree C, then it takes about 8 x 10^3 J to raise a liter by 2 degrees C.
extra heat ~= 2 x 10^15 L x 8 x 10^3 J/L
~= 1.6 x 10^19 JThen I compared that with annual human energy use:
share of annual human energy use
~= 1.6 x 10^19 J / 5.8 x 10^20 J/year
~= 2.7 x 10^-2
~= 2.7%So my first-pass answer was that the excess heat stored in just the top meter of the Mediterranean Sea is equivalent to about 2.7% of all human energy consumption worldwide in a year.
Calibration Score
Matt's Calibration Score: 90 / 100
Higher is better: earn points for accurate pegs, sound models, correct math, and a result close to the sourced answer. The image shows percent full of it: 100 minus the Calibration Score.
Pegs: 20/30. The area and world-energy pegs were close, with only small calibration updates needed.
Model: 30/30. Volume times density times specific heat times temperature change was the right model.
Math: 10/10. The arithmetic was clean.
Result: 30/30. The corrected answer closely matched the first pass.
Grounding facts
A square meter of ocean that is 1 meter deep contains about 1,000 kg of water. Warming that column by 2 degrees C takes:
heat per m2 per meter depth
~= 1,000 kg x 4.2 x 10^3 J/kg/C x 2 C
~= 8.4 x 10^6 JThat is about 2.3 kWh per square meter per meter of depth. The Mediterranean is roughly 2.5 x 10^12 square meters, so even a one-meter layer turns into a giant heat reservoir.
A 1-GW power plant running for one year produces about 3.2 x 10^16 J. The top-meter Mediterranean heat estimate is therefore hundreds of GW-years; a 10-meter layer is thousands of GW-years.
After checking sources
Check and recalibrate
Matt's top-meter calculation is very good for the model he chose. The Mediterranean area is usually given around 2.5 million km2, or 2.5 x 10^12 m2, so his area estimate was only about 20% low.
actual-ish area ~= 2.5 x 10^12 m2
top 1 m volume ~= 2.5 x 10^12 m3
top 1 m mass ~= 2.5 x 10^15 kgUsing water's specific heat, about 4.2 x 10^3 J/kg/degree C:
top 1 m heat ~= m c dT
~= 2.5 x 10^15 kg x 4.2 x 10^3 J/kg/C x 2 C
~= 2.1 x 10^19 JSo the top-meter answer is about 2 x 10^19 J, almost exactly Matt's order of magnitude.
The bigger judgment call is depth. Sea-surface temperature is not the whole ocean, but a summer surface mixed layer is often more than a single meter. If the warm anomaly represents the top 10 meters, multiply by 10:
top 10 m heat ~= 2.1 x 10^20 JIf the relevant layer is 20 meters, multiply again by 2:
top 20 m heat ~= 4.2 x 10^20 JIEA's recent global energy-demand peg is closer to 6.5 x 10^20 J/year than 5.8 x 10^20 J/year. That makes the comparison:
top 1 m share ~= 2.1 x 10^19 / 6.5 x 10^20
~= 3%
top 10 m share ~= 2.1 x 10^20 / 6.5 x 10^20
~= 30%
top 20 m share ~= 4.2 x 10^20 / 6.5 x 10^20
~= 65%A good calibrated answer is therefore about 2 x 10^19 J for the top meter, or about 2 x 10^20 J for a plausible shallow 10-meter layer. The result is not "humanity directly poured this much energy into the Mediterranean"; it is the amount of thermal energy stored in that water layer relative to a cooler counterfactual.
Post-check reflection
Matt's reflection
Looks like I was pretty close. My surface area estimate was a little low, sort of by accident. The worldwide energy-consumption peg also needs to be updated slightly, but it was close enough for this exercise.
That's a tremendous amount of energy, and this was a great way to help make the scope more easily appreciated.
Recommended memory peg
Remember water heat capacity ~= 4 x 10^3 J/kg/degree C. Also remember that 1 m3 of water ~= 1,000 kg, so 1 meter of ocean spread over 1 m2 has about 1,000 kg of water. For ocean heat content, use Q = m c dT.
Reader results
Bars show how submitted estimates sort into the answer choices from the gut-check prompt.