Abstract. We assessed the performance of ambient ozone (O3) and carbon dioxide (CO2) field calibration techniques when they were generated using data from one location and then applied to data collected at a new location. We also explored the sensitivity of these methods to the timing of field calibrations relative to deployments they are applied to. Employing data from a number of field deployments in Colorado and New Mexico that spanned several years, we tested and compared the performance of field-calibrated sensors using both linear models (LMs) and artificial neural networks (ANNs) for regression. Sampling sites covered urban, rural/peri-urban, and oil and gas production influenced environments. Generally, we found that the best performing model inputs and model type depended on circumstances associated with individual case studies. In agreement with findings from our previous study that was focused on data from a single location (Casey et al., 2017), ANNs remained more effective than LMs for a number of these case studies but there were some exceptions. In almost all cases the best CO2 models were ANNs that only included the NDIR CO2 sensor along with temperature and humidity. The performance of O3 models tended to be more sensitive to deployment location than to extrapolation in time while the performance of CO2 models tended to be more sensitive to extrapolation in time that to deployment location. The performance of O3 ANN models benefited from the inclusion of several secondary metal oxide type sensors as inputs in many cases.
Case Studies of Water Shut-Off Treatments in Oil and Gas Production Wells