Introducing new lightning schemes into the CHASER (MIROC) chemistry–climate model

Abstract

The formation of nitrogen oxides (NOx) associated with lightning activities (hereinafter designated as LNOx) is a major source of NOx. In fact, it is regarded as the dominant NOx source in the middle to upper troposphere. Therefore, improving the prediction accuracy of lightning and LNOx in chemical climate models is crucially important. This study implemented three new lightning schemes with the CHASER (MIROC) global chemical transport and climate model. The first lightning scheme is based on upward cloud ice flux (ICEFLUX scheme). The second one (the original ECMWF scheme), also adopted in the European Centre for Medium-Range Weather Forecasts (ECMWF) forecasting system, calculates lightning flash rates as a function of QR (a quantity intended to represent the charging rate of collisions between graupel and other types of hydrometeors inside the charge separation region), convective available potential energy (CAPE), and convective cloud-base height. For the original ECMWF scheme, by tuning the equations and adjustment factors for land and ocean, a new lightning scheme called the ECMWF-McCAUL scheme was also tested in CHASER. The ECMWF-McCAUL scheme calculates lightning flash rates as a function of CAPE and column precipitating ice. In the original version of CHASER (MIROC), lightning is initially parameterized with the widely used cloud-top height scheme (CTH scheme). Model evaluations with lightning observations conducted using the Lightning Imaging Sensor (LIS) and Optical Transient Detector (OTD) indicate that both the ICEFLUX and ECMWF schemes simulate the spatial distribution of lightning more accurately on a global scale than the CTH scheme does. The ECMWF-McCAUL scheme showed the highest prediction accuracy for the global distribution of lightning. Evaluation by atmospheric tomography (ATom) aircraft observations (NO) and tropospheric monitoring instrument (TROPOMI) satellite observations (NO2) shows that the newly implemented lightning schemes partially facilitated the reduction of model biases (NO and NO2), typically within the regions where LNOx is the major source of NOx, when compared to using the CTH scheme. Although the newly implemented lightning schemes have a minor effect on the tropospheric mean oxidation capacity compared to the CTH scheme, they led to marked changes in oxidation capacity in different regions of the troposphere. Historical trend analyses of flash and surface temperatures predicted using CHASER (2001–2020) show that lightning schemes predicted increasing trends of lightning or no significant trends, except for one case of the ICEFLUX scheme, which predicted a decreasing trend of lightning. The global lightning rates of increase during 2001–2020 predicted by the CTH scheme were 17.69 % ∘C−1 and 2.50 % ∘C−1, respectively, with and without meteorological nudging. The un-nudged runs also included the short-term surface warming but without the application of meteorological nudging. Furthermore, the ECMWF schemes predicted a larger increasing trend of lightning flash rates under the short-term surface warming by a factor of 4 (ECMWF-McCAUL scheme) and 5 (original ECMWF scheme) compared to the CTH scheme without nudging. In conclusion, the three new lightning schemes improved global lightning prediction in the CHASER model. However, further research is needed to assess the reproducibility of trends of lightning over longer periods.