144 



B. V. KOTLIAREVSKII 



where Sq^ and d^^ are the mean square errors in determining the increment 

 in gravity hetween two isoanomaUes, caused by the error in observations 

 and nondinearity of the field respectively. 



Comparing the relationships (6) and (7) with (4) and (5), it can be seen 

 that the required indices of accuracy E^ and D^ also split up into two 

 independent terms: 



F^ - F^ 



El 





(8) 

 (9) 



where 



•C-Om — 



fiOr 



F- — 



^Om = 



^0/ 



A> 



P 



&m 5 m P 



Thus, the problem of finding indices of absolute accuracy (e^ and d^) 



and the relative accuracy {E^ and 2)„^) of a gravimetric map is divided into 



two independent problems of finding the absolute (eo;„-. ^om) ^^^ ^^^^ relative 

 / g ^ 



jOm^ 1 errors due to the error in observations, and finding the abso- 



P 



-O/n 

 gm 



lute (f,-,^, (5,-^) and the relative I -i^, — — jerrors, caused by non-linearity 



\g'n PI 



of the field. 



Errors Due to Inaccuracies of the Observations (Com' ^om^ 



1. The derivation of the relationship £, 



Qm 



-^Oin 



(a, a). 



At the points x-^^ and x^ let g^ and g'g t»^ the true values of gravity, and 

 ^1 and g2 the observed values, and a^ and ^2 (Fig. 2) be the errors. 



-^ qU) 



I J — »— 



X, km 

 Fig. 2. 



