Author(s): Mesinger F (Mesinger, Fedor)

Title: Numerical methods in weather and climate modeling: From Milankovitch to the present, and beyond

Source: MILUTIN MILANKOVITCH ANNIVERSARY SYMPOSIUM: PALEOCLIMATE AND THE EARTH CLIMATE SYSTEM 110: 47-58

Date: 2005 

Document Type: Book in series : Proceedings Paper

DOI:  

Language: English

LCR: 0   CR: 44   LCS: 0   GCS: 0

Comment:  

Address: Serbian Acad Arts & Sci, Belgrade, Serbia Monteneg.

Reprint: Mesinger, F, Serbian Acad Arts & Sci, Belgrade, Serbia Monteneg.

E-mail:  

Author Keywords: numerical methods in weather and climate; history of numerical weather prediction; predictability; chaos; Eta model

KeyWords Plus: STEP-MOUNTAIN COORDINATE; ATMOSPHERIC MODELS; FLOW; REPRESENTATION; EQUATIONS; SCHEMES

Abstract: In this lecture I will recall some of the thoughts of Milutin Milankovitch (1879-1957; Milankovic, as written in Serbian, later on), one of the very few pioneers in using the tools of exact science to address the problem of weather and climate prediction. Before turning to climate, Milankovic considered the prediction of weather, only to conclude that "at least for the time being", using mathematics it was not doable. Two other great men of the time, Vilhelm Bjerknes and Lewis Fry Richardson, did go ahead, using very different approaches, and ending with just as different results. Milankovic, an engineer by training, and a person of an extraordinary common sense, was correct with his "for the time being": more knowledge, and also electronic computers, were needed for the first "real-time" numerical weather prediction to be achieved, at the time of this writing precisely 50 years ago. Steady progress followed, both in terms of "how to do it", and in understanding of "what might possibly be doable". The latter started with the earliest work on predictability by Phil Thompson, and was followed by the seminal work on "chaos" by Ed Lorenz. While we thus learned that there is a definite limit on atmospheric deterministic predictability, what will the weather be at a given place and time, moving toward that limit took and takes now just as well efforts of many people and numerous groups. In these efforts, numerical methods are just one of the requirements, but it is an essential one. The extraordinary diversity of ideas pursued testifies to the vitality of the subject, but also to the difficulty in identifying the benefit from one or another scheme or approach. I am convinced Milankovic would be pleased to know that-as I will review briefly-some of the ideas that have been proven useful have sprouted at his University of Belgrade; and that we are far ahead on the journey he largely started "through distant worlds and times", with good prospects of getting much farther still in the years to come.