Short answer is that models are unable to resolve highly localized boundary layer processes to a high degree of accuracy. Models resolution also hinders accuracy as coarse models will not replicate topographic or other terrestrial factors.

If I may add, once we try to get more accurate models with higher resolution and all, we also need more accurate and distributed measurements to help validate and calibrate the simulation model. So, there is a need for additional meteorological instrumentation, from satellites to distributed "weather stations" and then the need to synchronize and harmonize all these measurements with the proposed models.

One suggestion I would like to make: there is a lot of new installations of renewable energy (wind and solar, in particular) that embed good local weather monitoring. It should be mandatory (for instance, part of the authorization process to install such facilities) for these installations to share their raw weather measurements for better local, regional and even global weather forecast models.

Dear Guillermo and Lima,

To arrive at high resolution model and inclusion of sensitive weather feature like clouds and others; what about chaotic behavior of other external events that impact climate?

Regards

Oyedemi raises a very important point. Factors like resolution, the approximations in model parametrization or imperfect observations from which to start a climate model prediction all affect the realism of a model simulation but there is another much more fundamental limit to predictions.

Even with an absolutely perfect numerical model, the smallest of errors in initial conditions will result in the detailed daily weather being unpredictable after a few weeks. Similarly, again with an absolutely perfect model, the smallest of errors in the initial conditions will eventually grow so that after a couple of years into the simulation, climate variability like El Nino's will no longer be correctly predicted.

This means that we will never have perfect weather forecasts or climate predictions due to sensitivity to initial conditions. You can switch the argument around too, so that even with perfect obervations from which to start our predictions, any small error in the numerical model will again result in a fundamentally limited range for prediction skill.

However, not all is lost and we must also think about what can be predicted despite these limits: for weather forecasting it is still possible to predict the time average change in weather beyond the limit of daily predictability, so seasonal forecasts are possible for example. Similarly it is possible to predict the future change in the average global climate well beyond the limit of predictability for particular climate events like an El Ninoand so climate predictions are possible.

This is a deep and interesting topic and I would recommend reading Ed Lorenz's famous papers from 1963 and 1969 and also the many popular accounts of chaos that have been published since.

Best

Adam

Dear Adam,

Thanks for identify those limits and uncertainties in climate models. Improvements are on the way.

Regards

I agree with all the points raised earlier. My suggestion is to re-look into the efficiency of the input parameters being used in the models. For example, it has been demonstrated in the recent years that ocean heat content would yield better results compared to SST in cyclone and monsoon predictions. It is worth while to carry out similar exercise for climate models also. Besides, the efficiency of other parameters also should be explored.

Hi MM Ali,

Thanks for your contributions. Please can you highlight the parameters.

Regards

Hi Oyedemi,

Since I never worked in climate models I may not be able to give a satisfactory answer. For cyclones and monsoons, I attached two sample papers on ocean mean temperature (OMT) / ocean heat content (OHC) and cyclones/monsoons. They are simple and short papers. We have done some other studies which we can discuss, if you feel. The global OMT and OHC data are available at COAPS/FSU site. These data can be used in climate models to see the impact.

In case of cyclones and monsoons, modellers (both statistical and dynamical) take SST representing the heat energy available in the oceans. But SST is not a right parameter, particularly if you use satellite derived values. It represents only a few millimetre thick ocean whereas the actual heat energy is available in upper oceans. So we proposed (and even suggested a better approach to estimate) the use of OMT derived from altimeter observations. How this is done is given in a different paper. Similarly, LST we use is properly derived or we have to define a new metric using the available soil moisture, thus considering the heat energy stored over land?

Does the air temperature we take correctly represents the energy source or we have to get another parameter considering the available moisture? Are the models able to correctly represent the atmospheric heat energy by separately considering moisture and temperature or we have to get a new metric combining the two? From the crop canopy, particularly derived from the satellites, are we in a position to correctly represent the amount of solar energy reaching the ground?

These are some wild questions that come to my mind. A few of us may not agree to what I ask. They may be correct. So instead of discussing in public forum like can we directly correspond through emails? Mine is [email protected]/[email protected]. I repeat what I started in the beginning: I am not a modeller and I do not know the in and out of a model.

Dear MM Ali,

Fabulous information!

Regards

It is because weather is far too complex to be exactly predictable. This make the long-term average of weather, which we call climate even more difficult to predict. Those who say differently either don't know what they are tallking about, or they have interests to simplify the scientific truth.

Planet Earth is about 4.5 billion years old and in this long time there had been periods that were much, much warmer than it is today (e.g. during the time dinosaurs lived on our planet) and also periods that were far colder (ice ages). In other words: it would be strange when the climate would be the same indefinately. This never has been the case. There were also periods when CO2 concnetration was much higher than today. Indeed there was a time when no O2 was in the atmosphere... Climate change is something that bothers us, because we are effected, because we are there to be bothered. It does not effect Planet Earth in any meaningful way as the planet will be there even if Human aren't anymore (although also Planet Earth won't be eternal). Among many other challenges the scientific mainstream believes that the generation of special gases through the burning of fossil fuels is the driving force for climate change. A few, however, see human contribution little in global warming and some even talk of global cooling. In RG there have been many contributions that put global warming and global cooling as two different processes. What is different is the outcome of very complex processes. The outcome is either warming or cooling, and to make it even more complex both can happen - in different places.... or at differnt times (a specific place can have warmer summers and colder winters as a resukt of change of the climate....,). How much of this change is contributed by human activies we can discuss and do research about, but it would be strange if only humans have contributed to these changes, or if they would be entirely caused by natural processes - both is very unlikely the case. Sometimes people think that IPCC does investigate into global warming, but indeed it looks into climate change. The expression global warming is out since many years as it is acknowldeged that it is much more than around temperature change. Global cooling, if it exists is also climate change. In Germany we can have temperatures in winter of minus 20 degrees Celsius. That is winter and we call this winter weather. Some years winters are mild, other years very harsh, long and very cold. Weather and climate are different things, and the weather is never the same. It changes and also if the sun energy received on January 1 at a particular place would be the same every year (which is not the case as the sender (sun) sends not always the same) it still would result in different weather as the energy received is only one from many, many factors that makes the weather. In the discussion often people speak of a food crisis because of global warming. I doubt that this is the only possible interpretation. Today we produce considerable more food per capita than a few decades ago. When people are starving then it is usual because of the impacts of wars, civil wars, ethnic tensions, or just because of poverty, but not necessarily because of climate change, at least not so directly as often assumed. Here in the Pacific Island countries the talk of climate change and its impact on food security has been rather dominant, despite this part of the world does not know widespread undernutrition and when we look at malnutrition than we discover that the biggest concern is obesity, and when we would anlyze the causes then we would not find climate change, but MacDonalds and other fast food outlets. I think that a cooling effect (if it exists) due to reduced sun activity moderates anthropogenic warming impacts, the same other natural impacts on the climate would such as volcanic eruptions etc. We should realise that something as complex as climate has many contributing factors that lead to change. To assume that everything is caused by humans or everything by natural causes is far too simplistic. For that reason it is rather difficult to make any precise predictions of the future. So both (processes that contribute to warming and such contributing to cooling) not only can happen, but they do, and they always did. What now is sold as new ideas and change to previous scientific knowledge is actually nothing really new other than that one aspect from an extremely complex process is taken out and displayed as the whole truth.

Thank you for raising so many significant points. “Weather variables are too complex to be exactly predictable.”

The extinction of dinosaurs was due to unfavorable conditions of the planet attributed to meteors injections into the atmosphere, which cause the global temperature to warm. The ice age was another terrible state of the planet then as well. Notice also that in the last four solar cycles (1976- till date), solar flux and sunspot number have radically decreasing. The just concluded summer here was full of mixed weather (cold a lot of times, wet, and fewer days of heat waves).

The emission of CO2 in huge quantities was traceable to the advent of industrial revolution. There are a lot of observational evidences to support this fact even in this discussion on ResearchGate. I have no clue as per when CO2 concentrations were much higher that we have as at today.

There may be other forcing not yet incorporated and identified as the candidate for climate change, but I’m convinced that human activities on the planet including continuous and indiscriminately burning of fossil fuels and emission of greenhouse gases drive the climate system to change. These changes have significant consequences such as extreme events as noticeable across the globe recently.

It’s important to note that the metric to gauge the change in climate system is average temperature of the system. If the global mean temperature shifts to warming side, we have warming events; if it swings to the cooling phase, the earth experiences cooling effect.

The weather patterns cannot be the same. The natural forcings are dynamics, chaotic, and highly impulsive. The Sun determines energy budget of the climate. The Sun daily output is never the same.

Lastly, when people talk about food crisis; this simply means severe drought, flooding, erosion which abnormal affect crop yields. The sea foods seem to scare. The bees are also affected by planet conditions.

Three evidences are attached with this answer. The first figure is my own observation. In the West African region, I found terrestrial surface temperature shifts from cooling phase to warming phase around 1991. The result completely consistent with the time series of CO2 data (Figure 2) supplied by Kenneth in this RG. Figure 3 shows CO2 and methane changes with human population (also supplied by Kenneth).

Dear MM Ali,

Please visit the link below carefully:

https://climatedataguide.ucar.edu/climate-data

Regards

Dear all, it is impossible to have accurate and precise weather prediction models even climate prediction models due to the limitations in understanding of all physics involved in modification of weather or climatic events. In addition the already known physics are still difficulty to represent them fully in the prediction model but they are parameterized with lot of approximations. Resolution matter but understanding of all the physics and parameterization schemes matters the most.

It means we need comprehensive understanding of physics underlying climate change.

Yes, even the known physics are represented in the models with approximations. Do you know the closure problem in the atmospheric boundary layer?. To me we do not need the models to be accurate or precise in predicting weather or climate. Let us use them to guide the prediction of weather and climate. Note I said to guide

Hi Philbert,

Great answer!

Climate models are supposed to predict averages of weather and variability in this, not if it is going to rain next week. Climate models are doing this better and better even the predictions are not so that we like.

https://www.youtube.com/watch?v=PTJ28r9MFaU&feature=em-subs_digest

Hi Henrik,

Thanks for your contributions. and YouTube clip : Hansen: Earth is Back to Eemian Era w 6-9 METER Sea Level Rise #COP23

Regards

As long as the period of prediction is getting longer, the accuracy of prediction is getting more erroneous. That can be justified by the accumulation of rounding error, the definiency of the modeling algorithims, and data assimilation issues. Numerical prediction remains a difficult process.