Development of a quality-controlled chlorine gas concentration data base for the Jack Rabbit II field experiments

Abstract

The Jack Rabbit II (JR II) field experiments involved a series of trials in 2015 and 2016, each with 5–20 tons of pressurized liquefied chlorine released. All five trials in 2015 had vertically downward releases. The four trials in 2016 involved different release orientations. Roughly 200 chlorine sensors of four types (Jaz, Canary, MiniRAE, and ToxiRAE) were on the test grid at fixed locations during each trial, placed along 90 deg concentric arcs from 0.2 to 11 km from the source. These sensors use different technologies and have different detection ranges. All sensors were carefully calibrated, but pre- and post-trial challenge test results show that the MiniRAE sensors tended to drift. We applied special corrections to the MiniRAE data to correct for the drift. The roughly 19 million concentration records were further quality-assured and condensed to summary tables that include the maximum concentrations. This condensed data set was used in the model evaluation study described in the current special issue. Preliminary analyses suggest that, despite the necessary corrections to the MiniRAE data and the saturation of some sensors, the JR II concentration data in general are of good quality. For example, collocated sensors show good agreement, the arc-maximum normalized concentrations with downwind distance follow the −5/3 power law, and the limited sensor vertical profile data show that cloud heights increase with downwind distance. Future field campaigns of similar scale should carefully examine experimental setup and the latest detection technologies to avoid instrument saturation.