Monthly Archives: May 2007

A Short Summary Of Why Skillful Climate Prediction Is Much More Difficult Than Skillful Weather Prediction

Climate Science has already weblogged on the claim in the 2007 IPCC WG1 report that,

““Projecting changes in climate due to changes in greenhouse gases 50 years from now is a very different and much more easily solved problem than forecasting weather patterns just weeks from now. To put it another way, long-term variations brought about by changes in the composition of the atmosphere are much more predictable than individual weather events.â€? [from page 105]

This weblog provides a short summary of why such a claim is absurd.

First, all climate and weather models include two components; a dynamic core (which involves advection, the pressure gradient force, and the gravitational acceleration) and parameterized or prescribed) physical, chemical and biological processes. Only the dynamic core is basic physics. All parameterizations are engineering code which means they include tunable components.

Weather prediction models parameterize long- and short-wave radiative flux divergence, stable clouds and precipitation, deep cumulus clouds, turbulence, and air-sea and air-land fluxes. The state variables in weather model are the three components of velocity, temperature, pressure, density of air, and the three phases of water (and sometimes other gaseous and aerosol components). A detailed discussion of this type of model is given, for example, in

Pielke, R.A., Sr., 2002: Mesoscale meteorological modeling. 2nd Edition, Academic Press, San Diego, CA, 676 pp. [Table of Contents]

The state variables are initialized from real world observations such as from radiosonde and satellite data. If the weather model is a regional model, it obtains information through lateral boundary conditions. The dynamic core of the weather model, therefore, is constrained by the real-world initial conditions and lateral boundary conditions. Most of the surface boundary conditions are prescribed. This includes, for instance, sea surface temperature, sea ice coverage, vegetation, and snow cover. Only certain quantities, such as soil moisture and land surface temperature may be permitted to change in response to the land-air fluxes. When the initial conditions of the weather model are “forgotten”, the parameterizations must skillfully predict the evolution of the state variables from that time forward, which is the reason that the weather prediction accuracy degrades and becomes of no value after a certain time period (e.g. see).

A climate model, in contrast, must model more processes than in a weather model (such as biogeochemistry of vegetation on land and plants in the ocean; sea ice dynamics; aerosol processes; ocean circulation; ground freezing and thawing; snow accumulation and melt and sublimation, etc. – see). For some of these climate processes (which involve physics, biology and chemistry) they are modeled, as with a weather model, by a dynamical core and by parameterizations. These include sea ice dynamics and ocean circulation, which both have advection, pressure gradient and gravitational parts, as well as the parameterization of other effects (such as turbulence, phase changes of water). Some of the climate processes, such as biogeochemistry and biogeography have no dynamical core, and are completely parameterized models.

Thus, a climate model involves more parameterizations with their tunable components than for a weather model, as well as additional new state variables (such as salinity, ice, snow, vegetation type and its root depth etc) for which initial conditions are required for all of these variables.

The climate model also has no real world constraint such as supplied by real-world initial conditions (and for a regional model lateral boundary conditions). This real-world data constrains its predictions. Instead, the state variables required for the dynamic core of each component of the climate model (i.e. the state variables for the atmosphere, land, ocean and continental ice) must be generated from the parameterizations!

The claim by the IPCC that an imposed climate forcing (such as added atmospheric concentrations of CO2) can work through the parameterizations involved in the atmospheric, land, ocean and continental ice sheet components of the climate model to create skillful global and regional forecasts decades from now is a remarkable statement. That the IPCC states that this is a “much more easily solved problem than forecasting weather patterns just weeks from now” is clearly a ridiculous scientific claim. As compared with a weather model, with a multi-decadal climate model prediction there are more state variables, more parameterizations, and a lack of constraint from real-world observed values of the state variables.

Leave a comment

Filed under Climate Science Misconceptions

Guest Weblog by Barry H. Lynn, Richard Healy, and Len Druyan

Introduction by Roger A. Pielke Sr.

Climate Science has had a very productive e-mail exchange of perspectives in response to the weblog of May 14 2007. The authors of the article referred to in the weblog have graciously agreed to write a guest weblog which is given below. For background on the authors, a brief biographical summary of each scientist is:

Dr. Barry Lynn is a research scientist at the Hebrew University of Jerusalem. The research on climate change was conducted while he was an associate research scientist at Columbia University and Carnegie Mellon University. Dr. Lynn’s interests include studying the impacts of “greenhouse” gases on climate and the effect of aerosols on precipitation. Many of his papers have been published by the AMS and JGR. He is also the C.E.O of Weather It Is LTD (, a company that produces weather forecasts and climatological information, with an emphasis on deriving new economic applications from such products.

Rick Healy is a systems analyst at the National Ocean Sciences Accelerator Mass Spectrometry Facility (NOSAMS) at the Woods Hole Oceanographic Institution, developing computational methods in performing high precision radiocarbon analysis. He also collaborates with the NASA/Goddard Institute for Space Studies (GISS) climate-modeling group in New York. His interests include Regional Climate Impacts using integrated regional climate models to study climate change issues. He also collaborates with scientists at UMass Amherst in paleoclimatetracer studies using the GISS d18O tracer model.

Dr. Druyan is a Senior Research Scientist and the Director of the Center for Climate Systems Research which is a unit of the Earth Institute at Columbia University. Alternatively referred to as “GISS at Columbia”, CCSR is the administrative umbrella for many Columbia University research scientists based at the Goddard Institute for Space Studies. Dr. Druyan’s research interests focus on climate variability in tropical latitudes. His published work relates to a range of themes, including the Indian and West African summer monsoons, Sahel drought, African wave disturbances, climate change impacts on tropical cyclones, El Niño and other sea-surface temperature anomaly impacts on regional climates and seasonal climate prediction for Brazil. He has conducted climate simulation studies using several versions of the GISS GCM and more recently using a regional climate model (RM3) that represents variables at higher spatial resolution. Dr. Druyan’s research group is using the RM3 for collaborative research in the context of the African Monsoon Multidisciplinary Analysis (AMMA) and the West African Monsoon Modeling and Evaluation (WAMME).

The Guest Weblog follows:

This is a brief response to the posted critiques of our recent paper in the Journal of Climate. (Lynn et al., see ) We have since had a constructive dialogue with Dr. Pielke by email and we appreciate his giving us this opportunity for clarification on his blog. We hope to correct the mistaken impression that we were in any way looking to sensationalize dangers from global warming. Contrary to the impression promoted on the blog, we were diligent in our research. In addition, our paper was peer reviewed and it underwent revisions consistent with the suggestions of three (presumably) professional reviewing scientists.

The study was based on both observed data and model simulations. A significant result of the observational analysis was finding the strong inverse relationship between eastern U.S precipitation frequency and maximum surface temperatures (see figure). Dr. Pielke’s main criticism seems to focus on the poor performance of the AOGCM that provided data for driving our regional model. The AOGCM admittedly has deficiencies, as do all models. We believe that this AOGCM has comparable skill (or lack of skill) to other GCMs that formed the basis of the IPCC fourth assessment. The AOGCM is a tool, albeit imperfect, for projecting the broad scale climate consequences of increasing concentrations of greenhouse gases. It accounts for the distribution of oceans and continents and all of the major interactions between the different Earth systems affecting the climate. However, it has serious flaws regarding the simulation of regional climate.

Dr. Pielke maintains that dynamic downscaling by even the most skillful regional model cannot improve the simulation of a flawed climate simulation by a GCM. However, he concedes that the regional model solution is less sensitive to the driving GCM when the nested domain is large, as in our case. We also counter that (in our double-nested experiments) since the outer regional model domain extended from the Pacific to the Atlantic Oceans, no GCM data generated over the US was ever used to drive the regional model. In addition, we claim that crucial aspects of our climate simulations were indeed improved – by the alternative moist convection schemes operating at high horizontal resolution in the regional model. At least one version of our regional model simulated summertime precipitation frequencies that were much more realistic than the GCM.

Dr. Pielke correctly points out that the regional model cannot correct for GCM errors in the timing or trajectories of synoptic systems. We reply that radiation feedbacks that depend on the frequency of precipitation, mean ground wetness and frequency of cloudiness are far more important in determining rates of warming over the next 80 years. This downscaling produced a greater warming trend over the eastern US into the 2080s than the GCM because it did not make the mistake of “predictingâ€? rain on 65% of the summer days (see figure). Was this result adversely affected by the GCM data streaming in at the boundaries over the Pacific and Atlantic? The same GCM boundary conditions were used to drive another version of the regional model with a convection scheme that made the same mistake (as the GCM) of predicting rain too frequently. This version produced a more gradual warming trend just like the GCM. A third version that underestimated afternoon precipitation predicted the most severe warming trend. Based on all of this evidence, we are convinced that the radiation feedbacks created by the precipitation regime control the warming rates, and that our paper’s “apocalypticâ€? prediction of 5°C warming over the eastern US between the 1990s and 2080s is the most realistic prediction – a correction if you will to the underestimate of IPCC models that rain too frequently. See or

Relationship between the JJA anomalies of mean maximum T for the eastern US vs. the percent of rainy days in the corresponding seasons, 1977-2004.

Precipitation frequency (percent of rainy days during JJA) for JJA 1993-97 and JJA 2083-87 over the eastern US for observations and model versions. “Scaledâ€? observations refers to frequencies within 4° x 5° AOGCM grid elements.

Leave a comment

Filed under Guest Weblogs

More Presentation Of Climate Predictions as Scientific Fact

There is a new Science paper

Richard Seager, Mingfang Ting, Isaac Held, Yochanan Kushnir, Jian Lu, Gabriel Vecchi, Huei-Ping Huang, Nili Harnik, Ants Leetmaa, Ngar-Cheung Lau, Cuihua Li, Jennifer Velez, and Naomi Naik: Model Projections of an Imminent Transition to a More Arid Climate in Southwestern North America Published online 9 April 2007 [DOI: 10.1126/science.1139601] (in Science Express Reports) [thanks to Willie Soon for alerting us to it]

The abstract reads

“How anthropogenic climate change will impact hydroclimate in the arid regions of Southwestern North America has implications for the allocation of water resources and the course of regional development. Here we show that there is a broad consensus amongst climate models that this region will dry significantly in the 21st century and that the transition to a more arid climate should already be underway. If these models are correct, the levels of aridity of the recent multiyear drought, or the Dust Bowl and 1950s droughts, will, within the coming years to decades, become the new climatology of the American Southwest.”

An excerpt from the paper reads,

“In the multi-model ensemble mean there is a transition to a sustained drier climate that begins in the late 20th and early21st centuries”


“The drying of subtropical land areas that, according to the models is imminent or already underway, is unlike any climate state we have seen in the instrumental record. It is also distinct from the multidecadal megadroughts that afflicted the American Southwest during Medieval times …which have also been attributed to changes in tropical SSTs…The most severe future droughts will still occur during persistent La Niña events but they will be worse than any since the Medieval period because the La Niña conditions will be perturbing a base state that is drier than any experienced recently.”

This result appears to be contradictory to an earlier study

Trenberth, K. E., T. J. Hoar, El Niño and climate change, Geophys. Res. Lett., 24(23), 3057-3060, 10.1029/97GL03092, 1997.

Their abstract reads,

“A comprehensive statistical analysis of how an index of the Southern Oscillation changed from 1882 to 1995 was given by Trenberth and Hoar [1996], with a focus on the unusual nature of the 1990–1995 El Niño-Southern Oscillation (ENSO) warm event in the context of an observed trend for more El Niño and fewer La Niña events after the late 1970s. The conclusions of that study have been challenged by two studies which deal with only the part of our results pertaining to the length of runs of anomalies of one sign in the Southern Oscillation Index. They therefore neglect the essence of Trenberth and Hoar, which focussed on the magnitude of anomalies for certain periods and showed that anomalies during both the post-1976 and 1990–mid-1995 periods were highly unlikely given the previous record. With updated data through mid 1997, we have performed additional tests using a regression model with autoregressive-moving average (ARMA) errors that simultaneously estimates the appropriate ARMA model to fit the data and assesses the statistical significance of how unusual the two periods of interest are. The mean SOI for the post-1976 period is statistically different from the overall mean at <0.05% and so is the 1990–mid-1995 period. The recent evolution of ENSO, with a major new El Niño event underway in 1997, reinforces the evidence that the tendency for more El Niño and fewer La Niña events since the late 1970s is highly unusual and very unlikely to be accounted for solely by natural variability.”

The 2007 Science paper is yet another example of overselling of a process study as we discuss in our book

Cotton, W.R. and R.A. Pielke, 2007: Human impacts on weather and climate, Cambridge University Press, 330 pp.

The Seager et al 2007 paper is clearly is an example of the publication of a prediction, which has yet to be tested in its accuracy, as a scientific contribution. At least, with their claim of almost perpetual drought in the Southwest USA, we can track this over the next few years to either refute or support their conclusions.

Leave a comment

Filed under Climate Change Metrics, Climate Models

Summary Of Climate Science Perspective

I have decided to post the responses to several comments as a weblog since they succinctly capture The Climate Science’s perspective on the climate science issue [thanks to Tom, Logically Speaking, Allan J. and Frank K. for your constructive comments and contribution to the discussion on Climate Science!]:

1. Regarding the issue of the “butterfly effect”, all of the information can dissipate into heat without upscaling if it is a small enough perturbation. This was clearly explained by Professor Richard Ekyholt who is an international recognized expert on chaos and nonlinear dynamics. [see].

2. Climate prediction includes all aspects of weather prediction plus all other components of the climate system. On weather prediction time scales, many aspects of the climate system can be prescribed as constant in that time period; e.g. sea surface temperatures. On longer time scales, these components of the climate system must be predicted.

Thus nonlinearities in medium and long term feedbacks become important. We have discussed this issue in the multi-authored paper

Rial, J., R.A. Pielke Sr., M. Beniston, M. Claussen, J. Canadell, P. Cox, H. Held, N. de Noblet-Ducoudre, R. Prinn, J. Reynolds, and J.D. Salas, 2004: Nonlinearities, feedbacks and critical thresholds within the Earth’s climate system. Climatic Change, 65, 11-38.

The observations show nonlinear behavior on all time scales, with the IPCC models demonstrating absolutely no skill at predicting any of them.

What this means for predicting climate decades from now in response to the human input of CO2, or other human climate forcing, is that we do not know what the effect will be. It could move the climate system towards or away from important climate regime shifts.

The prudent path is to reduce the human forcing of the climate system since we do not know its consequences. However, what the IPCC has failed to do is to adequately assess the relative role of all climate forcings. As we have discussed on Climate Science and have published on (e.g. see and see), the heterogeneous climate forcings due to aerosols and land use/land cover change appear to be more significant in terms of our future climate then the radiative effect of CO2.

3. The approach, to assess vulnerabilities of important societal and economic resources, was recommended in the multi-authored book

Kabat, P., Claussen, M., Dirmeyer, P.A., J.H.C. Gash, L. Bravo de Guenni, M. Meybeck, R.A. Pielke Sr., C.J. Vorosmarty, R.W.A. Hutjes, and S. Lutkemeier, Editors, 2004: Vegetation, water, humans and the climate: A new perspective on an interactive system. Springer, Berlin, Global Change – The IGBP Series, 566 pp. {see Chapter E Section 3, Section 5, and Section 7, for example]

An important, much-needed perspective on this subject was recently published

Pielke, Jr., R.A., Prins, G., Rayner, S. and Sarewitz, D., 2007. Lifting the taboo on adaptation. Nature, Vol. 445, pp. 597-598.

The IPCC should have started their assessment by first assessing what are the important vulnerabilities to essential social and environmental resources. Instead, they chose to cloak energy policy in terms of an inappropriately narrow view of the climate system.

Leave a comment

Filed under Climate Change Forcings & Feedbacks, Vulnerability Paradigm

WG1 IPCC Chapter 1 – More Scientifically Erroneous Statements

Climate Science has selected two errors in Chapter 1 of the 2007 WG1 IPCC Report to highlight in this weblog.

These are

1. They write

“This is the so-called butterfly effect: a butterfly flapping its wings (or some other small phenomenon) in one place can, in principle, alter the subsequent weather pattern in a distant place. At the core of this effect is chaos theory, which deals with how small
changes in certain variables can cause apparent randomness in complex systems.” [page 105]

This is an incorrect statement of what the “butterfly effect” really means, as discussed on Climate Science

What is the Butterfly Effect?

More on the Butterfly Effect

The perpetuation of the incorrect understanding (that in the real climate system, such a small perturbation can affect large scale weather systems) illustrates how poorly written and researched the IPCC Chapter actually is.

2. The second error (and it is a big one) is their unsubstantiated claim that

“Projecting changes in climate due to changes in greenhouse gases 50 years from now is a very different and much more easily solved problem than forecasting weather patterns just weeks from now. To put it another way, long-term variations brought about by changes in the composition of the atmosphere are much more predictable than individual weather events.” [from page 105]

This is a remarkable claim, and forms the basis of the entire IPCC concept. The hypotheses that need to be tested to support their claim (and which should have been presented in Chapter 1 of the IPCC Report) are discussed on the Climate Science weblogs:

Are Multi-Decadal Global Climate Simulations Hypotheses? Have They Been Tested, and, If So, Have the Hypotheses As Represented By the Models, Been Falsified?

Three Hypotheses On The Role of Human-Climate Forcings In The Climate System

Comment on the Real Climate Post on “Short and Simple Arguments For Why Climate Can Be Predictedâ€? . Climate Science Disagrees With Their Statement

Is Climate Prediction Sensitive To Initial Conditions?

Their claim that�

Projecting changes in climate due to changes in greenhouse gases 50 years from now is a very different and much more easily solved problem than forecasting weather patterns just weeks from now.”

is such an absurd, scientifically unsupported claim, that the media and any scientists who swallow this conclusion are either blind to the scientific understanding of the climate system, or have other motives to promote the IPCC viewpoint. The absurdity of the IPCC claim should be obvious to anyone with common sense.

Leave a comment

Filed under Climate Science Misconceptions

The Need To Improve Land-surface Model Hydrology In Studying the Climate System

There is a excellent new paper which demonstrates an additional complexity for accurately modeling the climate system. It is

Gulden, L. E., E. Rosero, Z.-L. Yang, M. Rodell, C. S. Jackson, G.-Y. Niu, P. J.-F. Yeh, and J. Famiglietti (2007), Improving land-surface model hydrology: Is an explicit aquifer model better than a deeper soil profile?, Geophys. Res. Lett., 34, L09402, doi:10.1029/2007GL029804.

The abstract reads

“The interaction between groundwater and the atmosphere has a potentially significant influence on regional climate variability. As a result, researchers have focused increasing attention on improving how subsurface hydrology is represented within land-surface models, which are computer programs that simulate the exchange of surface water and energy fluxes between the land surface and the atmosphere. Using the National Center for Atmospheric Research’s Community Land Model over an area representing Illinois, Gulden et al. (2007) sought to determine how different representations of subsurface hydrology behaved under three different uncertainty scenarios: when everything about the system is known, when nothing about the system is known, and when limited knowledge about the system exists. They found that while the different representations can simulate known monthly values of Illinois’ average terrestrial water storage, a more physically realistic representation of subsurface processes reduces the sensitivity of the model to errors in model parameters. Further, approximate knowledge of model parameters is not sufficient to guarantee realistic model performance because interaction among parameters exacerbates errors.”

Among their conclusions is the answer to their posed question

“Does Knowledge of Parameter Ranges Guarantee Reasonable Model Output?”

which is

“Because of parameter interaction, knowledge of approximate parameter values is insufficient to guarantee realistic simulation of dTWS [total column terrestrial water storage].”


” In the foreseeable future, for large model domains, the scientific community is unlikely to be able to confidently assign subsurface hydrologic parameters either by direct observation or by calibration against subsurface hydrologic observations. Decreasing the sensitivity of model output to faulty parameter choices is therefore of utmost practical importance for improving model prediction capability. However, if soil texture properties are the dominant control on regional subsurface hydrologic variation in nature, then AQUIFER’s lower sensitivity to parameter values is likely problematic, and a significant increase in data collection and subsequent parameter estimation is warranted.”

Since this subsurface water is such an important part of the surface energy and water budget, the limited knowledge of this component of the climate system introduces another uncertainty in the ability to make skillful multi-decadal climate forecasts.

Leave a comment

Filed under Climate Change Forcings & Feedbacks

Request For Photographs Of GHCN sites – A Need For Documentation

Several excellent comments on Climate Science over the past week (see) and on another weblog (see) have emphasized the value of photographs of surface weather stations that are used to measure temperatures that are used in the construction of the land part of the multi-decadal global average surface temperature trends. These stations are referred to as being in the Global Historical Climate Network -GHCN (see and see). The United States component is referred to as the USHCN.

The World Meteorological Organization (WMO) and the National Climate Data Center (NCDC) in the USA, however, have not insisted that such photographs be taken. As we and others have shown in peer reviewed papers; e.g.

“The Geoprofile metadata, exposure of instruments, and measurement bias in climatic record revisitedâ€? by Rezaul Mahmood, Stuart A. Foster and David Logan June 30, 2006 International Journal of Climatology

“Land use/land cover change effects on temperature trends at U.S. Climateâ€? by R. C. Hale, K. P. Gallo, T. W. Owen, and T. R. Loveland June 3 2006 Geophysical Research Letters

Davey, C.A., and R.A. Pielke Sr., 2005: Microclimate exposures of surface-based weather stations – implications for the assessment of long-term temperature trends. Bull. Amer. Meteor. Soc., Vol. 86, No. 4, 497–504.

Pielke Sr., R.A. J. Nielsen-Gammon, C. Davey, J. Angel, O. Bliss, M. Cai, N. Doesken, S. Fall, D. Niyogi, K. Gallo, R. Hale, K.G. Hubbard, X. Lin, H. Li, and S. Raman, 2007: Documentation of uncertainties and biases associated with surface temperature measurement sites for climate change assessment. Bull. Amer. Meteor. Soc., in press.

this is a very serious issue, as these temperatures form the basis of the claims of the magnitude of global warming (see Figure SPM.3a in the 2007 IPCC SPM).

This weblog requests that readers of Climate Science and their colleagues photograph GHCN sites within their countries (using digital cameras, if possible) and post as comments on Climate Science, or send in an e-mail to us directly. All sites are needed (those with good exposure should also be photographed).

The following url will lead you to the station names. But you have to go country by country in order to see the station names [Thanks to Khishig Jamiyansharav for providing the links for this information!].

If we receive a sufficient number, we will collect the photograph and summarize by country in a report. The requested procedure to complete the photographs is to take one of the instrument facility itself, and then stand at the facility and take photographs towards the north, east, south and west. This is the method we used in our paper

Davey, C.A., and R.A. Pielke Sr., 2005: Microclimate exposures of surface-based weather stations – implications for the assessment of long-term temperature trends. Bull. Amer. Meteor. Soc., Vol. 86, No. 4, 497–504.

An even more detailed protocol to document the GHCN sites is given in the report

Jamiyansharav, K., D. Ojima, and R.A. Pielke Sr., 2006: Exposure characteristics of the Mongolian weather stations. Atmospheric Science Paper No. 779, Colorado State University, Fort Collins, CO 80523, 75 pp.

With enough of these photographs, we may be able to educate policymakers and scientists on the serious issue of poor, non-spatially representative exposure of the temperature measurements at many locations.

Leave a comment

Filed under Climate Change Metrics