--- title: "How the ocean remembers changing weather" description: "How can rapidly changing weather produce slowly changing climate?" canonical_url: "https://sandboxphysics.com/stories/thermal-complexity/hasselmann-weather-climate/" language: "en" page_type: "story" alternate_languages: zh-CN: "https://sandboxphysics.com/zh/stories/thermal-complexity/hasselmann-weather-climate/" last_modified: "2026-10-24" --- [Why climate responds slowly to a changing atmosphere](https://sandboxphysics.com/stories/thermal-complexity/#climate) Heat · Chance · Collective behaviour # How the ocean remembers changing weather How can rapidly changing weather produce slowly changing climate? Sandbox Physics · 2026-10-24 Nobel contribution: 2021-1 ## A short push can leave a long response A windy afternoon can stir the sea, change evaporation and alter how much heat leaves its surface. The wind may disappear overnight. The water it cooled takes longer to recover. Repeat such disturbances for years and the ocean carries a changing temperature history, even when no single storm lasts very long. Klaus Hasselmann made this difference in response times central to his 1976 theory of stochastic climate models. “Stochastic” means that random variations are included explicitly. Working at the Max Planck Institute for Meteorology in Hamburg, he asked what happens when fast weather keeps pushing a slower climate system. Some slow climate variability could arise from the accumulated response to those short disturbances. Sources: [2] [5] ## Random disturbances do not cancel immediately Imagine briefly warming or cooling a large tank at irregular times. The average heat input might be zero, but every burst of heating need not be followed immediately by equal cooling. A run of disturbances in the same direction leaves the tank warmer or colder for a while. Its temperature remembers their sequence. Heat capacity controls how strongly each disturbance changes temperature. A large heat capacity spreads a given heat input over a smaller temperature change. The long persistence of ocean temperatures comes from thermal inertia together with the rates of heat loss and exchange. A slow system weakens rapid variations in temperature, so longer variations stand out. If there were no restoring tendency, the spread of possible temperature histories would keep growing. Hasselmann’s theory included stabilising feedbacks: when a warm anomaly loses extra heat, the tendency to return provides a finite memory. A spectrum shows how much variability comes from changes on each time scale. Fast forcing and slow relaxation together can put more of it at the slow end, often called red noise. Sources: [2] [3] ## The next question was where to look in the ocean Claude Frankignoul and Hasselmann developed the next paper in the series in 1977. They applied the stochastic approach to sea-surface temperature anomalies and changes near the thermocline, the layer where ocean temperature changes rapidly with depth. This brought the argument closer to quantities oceanographers could measure. A useful comparison asks how large temperature variations are at different time scales and how quickly their memory fades. Two ocean records can have the same average and still differ strongly in persistence. Their time dependence can reveal which combination of random input, heat storage and restoring is plausible. Matching this structure tests more than whether a line happens to pass through a few warm years. Sources: [3] ## A warming signal has to stand out against this background Natural variability matters when asking whether an externally forced change has appeared. Hasselmann later developed fingerprint methods that compare a predicted pattern of change with observations, allowing for the size and correlations of natural fluctuations. A pattern is more informative where the background varies little; strongly correlated observations cannot be treated as independent repetitions. His 1993 work formulated an optimal linear filter for a specified signal and an estimated variability covariance. Here covariance describes which measurements tend to fluctuate together. The method requires both a physical signal pattern and a credible description of background variability. Detection and attribution grew through work by many collaborators, including Gabriele Hegerl and Ben Santer. The 2021 Nobel Prize paired Hasselmann’s contribution with Manabe’s climate modelling. One route followed how added greenhouse gases change temperature; the other clarified the variability against which such changes are judged. Together, these approaches linked a predicted temperature response to a statistical test against the climate’s own variability. Sources: [4] [5] [1] Continue with the physics ## Try the related models ### [Compare fast and slow temperature responses](https://sandboxphysics.com/labs/climate-energy-balance/) Increase ocean exchange and compare surface and deep-ocean warming. The deterministic two-box model shows heat storage and relaxation under prescribed forcing; Hasselmann’s random weather forcing is explained in the reading. ## Sources and reading notes 1. [Royal Swedish Academy of Sciences (2021), Physics prize announcement](https://www.nobelprize.org/prizes/physics/2021/press-release/) The official announcement separates the shared Manabe–Hasselmann climate contribution from Parisi’s disorder contribution. 2. [Hasselmann (1976), Stochastic climate models, Part I: Theory](https://doi.org/10.3402/tellusa.v28i6.11316) Read the fast weather/slow climate split, the random forcing term and stabilising feedback. Slow variability can arise without a slowly varying external driver. 3. [Frankignoul and Hasselmann (1977), Stochastic climate models, Part II](https://web.lmd.jussieu.fr/~jyg/frankignoul_hasselmann-stochastic_climate_models_SST_anomalies-tellus-1977.pdf) Read how the stochastic idea is applied to sea-surface temperature anomalies and thermocline variability, with spectra and persistence as observables. 4. [Hasselmann (1993), Optimal Fingerprints for the Detection of Time-dependent Climate Change](https://doi.org/10.1175/1520-0442(1993%29006%3C1957:OFFTDO%3E2.0.CO;2) The specified signal and the estimated covariance of natural variability enter the optimal linear detector together. This is a detection method, distinct from the 1976 variability theory. 5. [Max Planck Institute for Meteorology, The Nobel Prize in Physics 2021](https://mpimet.mpg.de/en/institute/the-nobel-prize-in-physics-2021) The institute traces Hasselmann’s work and identifies the collaborators behind later detection and attribution studies. [Heat, chance and collective behaviour](https://sandboxphysics.com/stories/thermal-complexity/) [Manabe and Wetherald followed heat through the atmosphere](https://sandboxphysics.com/stories/thermal-complexity/manabe-radiative-convective/)