Performance Assessment of the Optical Transient Detector and Lightning Imaging Sensor. Part I: Predicted Diurnal Variability

Abstract

Laboratory calibration and observed background radiance data are used to determine the effective sensitivities of the Optical Transient Detector (OTD) and Lightning Imaging Sensor (LIS), as functions of local hour and pixel location within the instrument arrays. The effective LIS thresholds, expressed as radiances emitted normal to cloud top, are 4.0 ± 0.7 and 7.6 ± 3.3 ?J sr?1 m?2 for night and local noon; the OTD thresholds are 11.7 ± 2.2 and 16.8 ± 4.6 ?J sr?1 m?2. LIS and OTD minimum signal-to-noise ratios occur from 0800 to 1600 local time, and attain values of 10 ± 2 and 20 ± 3, respectively. False alarm rate due to instrument noise yields ?5 false triggers per month for LIS, and is negligible for OTD. Flash detection efficiency, based on prior optical pulse sensor measurements, is predicted to be 93 ± 4% and 73 ± 11% for LIS night and noon; 56 ± 7% and 44 ± 9% for OTD night and noon, corresponding to a 12%?20% diurnal variability and LIS:OTD ratio of 1.7. Use of the weighted daily mean detection efficiency (i.e., not controlling for local hour) corresponds to σ = 8%?9% uncertainty. These are likely overestimates of actual flash detection efficiency due to differences in pixel ground field of view across the instrument arrays that are not accounted for in the validation optical pulse sensor data.