Dear Sebastiano, I´m not sure but I think the work You are looking for or at least a related one can be found following this link:

http://onlinelibrary.wiley.com/doi/10.1002/qj.49709239320/abstract

otherwise You may borrow a copy from some of the libraries listed in this other link

http://www.worldcat.org/title/objective-analysis-of-meteorological-fields/oclc/4629117

Hope it be useful. . 

Best regards

Alberto

Dear Sebastiano,

I have this reference which seems to be the original one (translated)

Gandin, L.: Objective analysis of meteorological fields (Gridrometeorol Izdat, Lenningrad (translation published by the Israel Program for Scientific Translation, Jerusalem, 1965), 242 pp., 1963

However, if you are interested in modern use of objective analysis, I suggest to you the following:

Rutherford I (1972) Data assimilation by statistical interpolation of forecast error fields. J Atmos Sci 29:809-815

Daley R (1991) Atmospheric data analysis. Cambridge University Press

Robichaud A, Ménard, R (2014a) Multi-year objective analyses of warm season ground-level ozone and PM2.5 over North America using real-time observations and Canadian operational air quality models. Atmos Chem Phys 14:1-32. doi:10.5194Éacp-14-1-2014

Regards

Alain

Dear Alberto and Alain

Thank you very much for your replies, I surely will give a look to the cited references. I'm curious to give a look to the original work (Alain, the one you are citing) of Gandin because of I would like to see by myself how much the work is similar to Matheron's one.

Sincerely,

Sebastiano

Matheron was aware of Gandin's work and in at least one of his publications mentions this.  The underlying mathematics is basically the same for Matheron's work and for Gandin but there are some differences. Matheron emphasized the use of the variogram (a weaker form of stationarity) whereas Gandin stressed the use of the covariance function. Matheron was originally motivated by problems in ore reserve estimation (mining) whereas Gandin was interested in meteorology, the characteristics of the data for these two different applications influenced how they approached the problems. Look at some of the work of Grace Wahba in connection with Objective Analysis. Note that Matern was working at about the same time and his work overlaps with that of both Matheron and Gandin. All three did their work outside the English language literature hence it was not much known until later when translations appeared (Ganbin's work became known better when he immigrated to Isreal and his material was translated, Matheron's work became better known after the NATO ASI on Advanced Geostatistics in the Mining Industry (D. Riedel, 1976) and then when his student A.E. Journel moved to Stanford and published "Mining Geostatistics" with Ch. Huijbregts,  Academic Press (1978). Matern's work became better known when Springer published an English version (originally in Swedish) in 1986.  Matheron published a seminal theoretical paper (in English) in the journal "Advances in Applied Probability" in 1973

As an aside, "data assimilation" is a very important aspect of Gandin's work (also Grace Wahba) but not so much in Matheron's work.  Matheron and his group had early collaboration with Shell oil and they produced the software package "Bluepack" written for the DEC VAX (1978). Matheron also had early collaboration with people in hydrology.

See "Geostatistics", 2nd ed,  J.P. Chiles and P. Delfiner, J. Wiley. Both Chiles and Delfiner were in Matheron's original group.

Early on Matheron's work was not so well known in the statistics community/literature.  However a number of statisticians spent time at Fontainebleau but the numbers were always small because they lacked the space to accommodate them.

Very interesting Donald.

Thanks for increasing our knowledge about the subject.

Thank you very much Prof. Donald Myers!

And sorry for the late reply....

It is very interesting to analyse as these ideas developed and mixed....and it could be also interesting to understand how and if some of the basic ideas are evaluated in new approaches such as in machine learning (I'm thinking to support vector machines).

It would be very nice to have a kind of unifying book that highlight links and analogies between different approaches in spatial statistics and, probably, also in time series analysis. 

Sebastiano

Please note that the Wiley reference is to a review of the book in QJRMS rather than the work itself.

his work is fairly well known in the atmospheric sciences/meteorology literature but is essentially superceded by geostatistics which is far better known in many disciplines and multiple literature. G. Matheron was aware of Gandin's work.

The original translation of Gandin's work occurred after he moved to Isreal

You might want to look at "Spatial Variation" by B. Matern (Springer), an English translation from Swedish.

Both Gandin and Matern focus only on covariance functions (as opposed to variograms). Geostatistics has expanded enormously since Gandin's and Matern"s work appeared (both about 1965 but not translated until much later).