Exploring the relationship between surface PM_2.5 and meteorology in Northern India

Abstract

Northern India (23–31° N, 68–90° E) is one of the most densely populated and polluted regions in world. Accurately modeling pollution in the region is difficult due to the extreme conditions with respect to emissions, meteorology, and topography, but it is paramount in order to understand how future changes in emissions and climate may alter the region's pollution regime. We evaluate the ability of a developmental version of the new-generation NOAA GFDL Atmospheric Model, version 4 (AM4) to simulate observed wintertime fine particulate matter (PM2.5) and its relationship to meteorology over Northern India. We compare two simulations of GFDL-AM4 nudged to observed meteorology for the period 1980–2016 driven by pollutant emissions from two global inventories developed in support of the Coupled Model Intercomparison Project Phases 5 (CMIP5) and 6 (CMIP6), and compare results with ground-based observations from India's Central Pollution Control Board (CPCB) for the period 1 October 2015–31 March 2016. Overall, our results indicate that the simulation with CMIP6 emissions produces improved concentrations of pollutants over the region relative to the CMIP5-driven simulation. While the particulate concentrations simulated by AM4 are biased low overall, the model generally simulates the magnitude and daily variability of observed total PM2.5. Nitrate and organic matter are the primary components of PM2.5 over Northern India in the model. On the basis of correlations of the individual model components with total observed PM2.5 and correlations between the two simulations, meteorology is the primary driver of daily variability. The model correctly reproduces the shape and magnitude of the seasonal cycle of PM2.5, but the simulated diurnal cycle misses the early evening rise and secondary maximum found in the observations. Observed PM2.5 abundances are by far the highest within the densely populated Indo-Gangetic Plain, where they are closely related to boundary layer meteorology, specifically relative humidity, wind speed, boundary layer height, and inversion strength. The GFDL AM4 model reproduces the overall observed pollution gradient over Northern India as well as the strength of the meteorology–PM2.5 relationship in most locations.