EVALUATING THE ACCURACY OF A GRAVIMETRIC SURVEY 165 



to the value of this error. After analysing the errors, it is useful to consider 

 the errors introduced by the classical criteria and to compare them with the 

 criteria E^ and D^ suggested by the author. 



The value a gives the objective error in accuracy for the whole run of 

 a survey, including both the gravimetric and geodesic measurements, the 

 processing of the observations, the introduction of corrections, etc. However, 

 the value of this error does not make it possible to draw conclusions on the 

 accuracy of the observed field e^^ which, depending on the network density 

 and the character of the field, can have higher or lower values than a. Thus, 

 the "cr criterion" is essential but not sufficient to evaluate the quality of 

 the observed gravity field. 



It follows from the preceding that the evaluation of the accuracy of the 

 field from the value a is replaced by the evaluation from the value e^ or 

 E^, thus bearing in mind the parameters of the survey a and a, and also 

 the character of the field gQ and Iq. 



The second criterion, which is determined by the inequality p ^ 2.5 a, 

 does not take into account either the network density or the character of 

 the field, and is not adequate for either of these reasons. We have replaced 

 it by the criterion d^ —D^p. In particular, for the recommended value 

 Dj^ = 30%, we obtain p = 3.3 (5,„. This condition means that the map 

 of the gravity isoanomalies will have practically no sections where the 

 error in the gravity increment between two neighbouring isoanomalies would 

 be equal to the cross-section of the isoanomalies. 



Although a preKminary test of the proposed method with theoretical 

 examples and with actual material gave sufficiently reliable results, in later 

 practical use of this method, in some cases difficulties might be encountered 

 and individual faults discovered, which have not been mentioned by the 

 author. The future problems will be to check this method and for this pur- 

 pose the cooperation of production workers will be required. 



The author thanks K. E. Veselov, A. M. Lozinskii, L. V. Petrov 

 and N. N. Puzyrev for valuable advice and comments made during a dis- 

 cussion of the present article. 



REFERENCES 



1. B. A. Andreev, Calculation of a Network of Observations in Gravimetric Work. 



Documents of Ts. NIGRI. Geofizika, CoUection 5, ONTI NKTP SSSR, (1938). 



2. Yu. D. BuLANZHE, Some problems in the method of gravimetric work. Appl. Geo- 



physics, No. 1, Gostoptekhizdat, (1945). 



3. R. F. VoLODARSKli, Some problems of a gravimetric survey. Survey and Industrial Geo- 



physics, No. 11. Gostoptekhizdat, (1954). 



