The objective of this study is to present a method to reduce the setup error inherent in clinical depth dose measurements and, in doing so, to improve entrance dosimetry measurement reliability. Ionization chamber ͑IC͒ depth dose measurements are acquired with the depth scan extended into the air above the water surface. An inflection region is obtained in each resulting percent depth ionization ͑PDI͒ curve that can be matched against other measurements or to an inflection region obtained from an analogous Monte Carlo ͑MC͒ simulation. Measurements are made with various field sizes for the 6 and 18 MV photon beams, with and without a Pb foil in the beam, to determine the sensitivity of the dose inflection region to the beam conditions. The offset between reference and test data set inflection regions is quantified using two separate methods. When comparing sets of measured data, maxima in the second derivative of ionization are compared. When comparing measured data to MC simulation, the offset that minimizes the sum of squared differences between the reference and test curves in the ionization inflection region is found. These methods can be used to quantify the offset between an initial setup ͑test͒ position and the true surface ͑reference͒ position. The ionization inflection location is found to be insensitive to changes in field size, electron contamination, and beam energy. Data from a single reference condition should be sufficient to identify the surface location. The method of determining IC offsets is general and should be applicable to any IC and other radiation sources. The measurement method could reduce the time and effort required in the initial IC setup at a water surface as setup errors can be corrected offline. Given a reliable set of reference data to compare with, this method could increase the ability of quality assurance ͑QA͒ measurements to detect discrepancies in beam output as opposed to discrepancies in IC localization. Application of the measurement method standardizes the procedure for localizing cylindrical ICs at a water surface and thereby improves the reliability of measurements taken with these devices at all depths.

The depth-ionization gradient has a discontinuity and is maximized when the outer-radius of a submerged IC coincides with the water surface. This feature can be used to auto-align ICs to the water surface at the time of scanning and/or be applied retrospectively to scan data to quantify absolute IC depth. Utilization of the gCAP should yield accurate and reproducible depth calibration for clinical depth-ionization measurements between setups and between users.

The water-to-air scanning method coupled with gCAP analysis identifies the alignment of the IC OR to the water surface within the scanning resolution for all ICs studied. The gCAP method can precisely and reproducibly align the physical center of a given cylindrical IC with the water surface, be applied prospectively or retrospectively, and provides the prospect for automated water surface identification for scanning systems. The gCAP method eliminates the visual subjectivity inherent to current IC-to-water surface alignment techniques, has been validated with a wide variety of commercially available ICs, and should be independent of the scanning system used for data acquisition.