Why weather prediction got brilliant – but not climate predictions
Dr. Peter Ridd, a long-time researcher of the Great Barrier Reef and former academic at James Cook University, offers a perspective on the evolving capabilities of weather forecasting and the ongoing debate surrounding climate models.
Drawing on decades of experience in physics and environmental science, he highlights a crucial distinction: while weather prediction has advanced dramatically, climate modelling presents a fundamentally different and more uncertain challenge.
At the heart of modern weather forecasting lies physics—specifically the laws governing motion, thermodynamics, and radiation. These models simulate how air moves, how heat transfers, and how moisture behaves in the atmosphere. Central to this is Newton’s second law (force equals mass times acceleration), which allows scientists to calculate how air masses accelerate and change direction over time. By dividing the atmosphere into a three-dimensional grid and applying these equations, meteorologists can simulate wind patterns, temperature shifts, and pressure changes.
However, the key to accurate weather prediction is not just the equations—it is the starting point. Weather systems are highly sensitive to what physicists call initial conditions. To predict tomorrow’s weather, one must know today’s conditions with high precision: wind speed, humidity, pressure, and temperature across the globe. Even tiny inaccuracies can amplify over time, leading to vastly different outcomes—a phenomenon often associated with chaotic systems.
watch the video below:
Ridd argues that the major breakthrough in weather forecasting has not been a dramatic improvement in physical theory, but rather a revolution in data collection. Satellites now measure atmospheric conditions globally, including temperature and humidity profiles using infrared and microwave signals. Traditional weather balloons still play a role, but are limited in coverage. More recently, even GPS signals have been harnessed to infer atmospheric humidity, based on how signal timing is affected by water vapor.
These advances mean that meteorologists now have a far more accurate picture of the atmosphere at any given moment. Combined with powerful supercomputers and sophisticated statistical methods—similar in principle to how AI systems learn from past data—this has led to significant improvements in short-term forecasts. Events like cyclone paths, once highly uncertain, can now be predicted with remarkable accuracy.
Climate models, however, operate differently. Rather than predicting specific conditions on a given day, they aim to estimate long-term averages—typically over decades. Instead of starting from precise current conditions, they simulate how the climate system responds to changes in variables such as greenhouse gas concentrations.
According to Ridd, this difference has important implications. Because climate models focus on long-term trends rather than immediate states, they do not benefit as directly from improved measurement of current conditions. While weather models are continually refined with real-time data, climate models rely more heavily on assumptions, parameterisations (especially for complex processes like cloud formation), and large-scale approximations.
He also points out that different climate models can produce varying results, sometimes differing by several degrees in their projections. This divergence raises questions about uncertainty and model reliability, particularly when projections extend many decades into the future. Furthermore, validating these models is inherently difficult—by the time predictions can be fully tested, it may be far in the future.
Ridd’s broader message is not to dismiss modelling altogether, but to distinguish clearly between the demonstrated success of weather forecasting and the more uncertain domain of long-term climate prediction. He emphasises that while meteorology has benefited enormously from improved data and measurement technologies, climate science faces deeper challenges tied to complexity, scale, and verification.
In closing, he suggests that public discourse should recognise the achievements of weather forecasting professionals, whose work has become increasingly reliable. At the same time, he encourages continued scrutiny and open discussion around the assumptions and limitations of climate models—particularly given their influence on long-term policy decisions.
source: www.youtube.com
Tom
| #
Weather predictions brilliant? I must have missed that somehow.
Reply
Terry Shipman
| #
A number of years ago Tom Bonner, chief meteorologist for channel 4 in Little Rock Arkansas, made a snow prediction on the 6 PM newscast of little or no accumulation. When he came to work for the 10 PM newscast the station personnel and cameras were in the station parking lot waiting on him. They proceeded to pelt him with handfuls of “little or no accumulation.” About six inches worth. He was a much-beloved weatherman in Little Rock but he spent years living down that blown forecast.
Reply
Herb Rose
| #
All of Newton’s laws are wrong. Nothing in the universe travels in a straight line because the universe is formed by pi. If a satellite traveled in a straight line it would no longer be in equilibrium with the energy source it is in and all matter equalizes with its energy source.
Force does not equal mass times acceleration. A force is energy being transferred to a mass. That energy not only causes acceleration, or an increase in kinetic energy, but also increases the internal energy of the mass causing radiation of energy in other spectrums.
Gravity is not a function of mass as Galileo experiments showed and Kepler’s equation states.
Meteorologists do not what the thermometer or barometer are recording, have no understanding of the nature of water in the atmosphere or how energy flows so their understanding of the weather is the same as an astrologists understanding of how the universe works.
Reply