Anna, Ohio, at the junction of Cincinnati, Findlay and Kankakee arches, has been the site of repeated and sometimes damaging earthquakes. As part of a seismicity investigation near-surface (0.15-1 m depth) overcoring strain measurements were conducted in four limestone and dolomite quarries surrounding Anna, the results of which suggest a complicated regional stress pattern, with changes in the maximum horizontal stress direction from N45ºE to N15ºW. Since such abrupt stress fluctuations are rather atypical in the Midwest and since the reliability of near-surface tests in defining regional stress had not been established, we decided to remeasure the stresses at the same locations but at greater depths (0-200 m) using the hydrofracturing method. Owing to the weak horizontal bedding planes of the Silurian shaly limestones and shaly dolomites and the relatively shallow depths, many of the tests yielded horizontal fractures, providing information only on the vertical principal stress (-0.029 MPa/m depth). However, sufficient tests in each of three testholes resulted in vertical hydrofractures to enable the calculation of all three principal stresses. Within 50-170 m depth the horizontal stresses in all three quarries appear to belong to a uniform regional stress field defined by the following relations: sHmin = 5.1 + 0.014 D; sHmax= 10.1 + 0.014 D, where D is depth in meters. The direction of sHmax within 50-170 m depth is uniform at N70ºE (±15º). The hydrofracturing results obtained at Anna, Ohio are in agreement with the prevailing stress regime in the midcontinent both with respect to magnitudes and directions. The results suggest a possible thrust mechanism for a proposed NW-SE fault along which most of the earthquake epicenters have been located, and a left lateral strike-slip at greater depths. The focal mechanism solution of a recent major earthquake near Sharpsburg, Kentucky, some 200 km south of Anna, also indicates strike slip induced by a N60ºE trending sHmax. The discrepancies between the near-surface overcoring and the deeper hydrofracturing are attributed to a detachment or decoupling of the stress field at about 50 m depth which renders any shallow measurements unrepresentative of the regional state of stress.
In a recent paper (Haimson, 1981) it was shown that in six case histories stress magnitudes and directions as determined by hydro fracturing agreed surprisingly well with those resulting from overcoring tests when both methods were employed at comparable depths. The study suggests that both methods measure the same quantity, namely the stress tensor, and that either can be used with some confidence provided the tests are correctly carried out. However, the conclusions of the paper or of any other study do not necessarily justify measuring the stress by either method at one depth in order to project its value to a different depth. For example, at Waterloo, Wisconsin, hydrofracturing tests in two adjacent drill holes in Precambrian quartzite revealed that the stresses in the upper 40 m were considerably different in both magnitudes and directions from the stress regime in the 50 m to 270 m depth (Haimson, 1980).