Using machine learning to derive cloud condensation nuclei number concentrations from commonly available measurements

Abstract

Abstract. Cloud condensation nuclei (CCN) number concentrations are an important aspect of aerosol–cloud interactions and the subsequent climate effects; however, their measurements are very limited. We use a machine learning tool, random decision forests, to develop a random forest regression model (RFRM) to derive CCN at 0.4 % supersaturation ([CCN0.4]) from commonly available measurements. The RFRM is trained on the long-term simulations in a global size-resolved particle microphysics model. Using atmospheric state and composition variables as predictors, through associations of their variabilities, the RFRM is able to learn the underlying dependence of [CCN0.4] on these predictors, which are as follows: eight fractions of PM2.5 (NH4, SO4, NO3, secondary organic aerosol (SOA), black carbon (BC), primary organic carbon (POC), dust, and salt), seven gaseous species (NOx, NH3, O3, SO2, OH, isoprene, and monoterpene), and four meteorological variables (temperature (T), relative humidity (RH), precipitation, and solar radiation). The RFRM is highly robust: it has a median mean fractional bias (MFB) of 4.4 % with ≈96.33 % of the derived [CCN0.4] within a good agreement range of -60%<MFB<+60% and strong correlation of Kendall's τ coefficient ≈0.88. The RFRM demonstrates its robustness over 4 orders of magnitude of [CCN0.4] over varying spatial (such as continental to oceanic, clean to polluted, and near-surface to upper troposphere) and temporal (from the hourly to the decadal) scales. At the Atmospheric Radiation Measurement Southern Great Plains observatory (ARM SGP) in Lamont, Oklahoma, United States, long-term measurements for PM2.5 speciation (NH4, SO4, NO3, and organic carbon (OC)), NOx, O3, SO2, T, and RH, as well as [CCN0.4] are available. We modify, optimize, and retrain the developed RFRM to make predictions from 19 to 9 of these available predictors. This retrained RFRM (RFRM-ShortVars) shows a reduction in performance due to the unavailability and sparsity of measurements (predictors); it captures the [CCN0.4] variability and magnitude at SGP with ≈67.02 % of the derived values in the good agreement range. This work shows the potential of using the more commonly available measurements of PM2.5 speciation to alleviate the sparsity of CCN number concentrations' measurements.

Machine learning uncovers aerosol size information from chemistry and meteorology to quantify potential cloud-forming particles

“…(2) atmospheric state and composition variables has been developed and is described in detail elsewhere 33 . The present analysis focuses on [CCN0.4] for the purpose of demonstration and in future work will be extensible for the full CCN spectrum.…”

“…Here, the RFRM is trained on 30-year simulations by GEOS-Chem-APM: a state-of-the-science chemical transport model with detailed size-resolved microphysics. The present study uses the RFRM-ShortVars configuration 33 , a fast implementation 49 of Random Forest models 50 in the statistical computing language R 51 , retrained using PM 1 speciation variables as predictors in the absence of airborne measurements of PM A Linear Regression model with minimization of least squares using the fast column-pivoted QR decomposition method is also developed on the airborne dataset. This intentional overfitting is to obtain an effective representation of linear regression-like current GCMs' prescriptions of aerosol mass to number.…”

“…Machine learning (ML) is a subset of Artificial Intelligence (AI) where computers are trained on a large number of scenarios to acquire knowledge by statistical learning and without explicit instructions. While ML has been in use (the humble and ubiquitous linear regression model) for the last several decades 21,22 , in recent years, novel techniques and rapid advances in this field have led to its emergent applications in the atmospheric sciences [23][24][25][26][27][28][29][30][31][32][33] , especially in grappling with ordinal, non-linear, complex, and massive amounts of data. It is key, however, that these increasingly black-box ML/AI techniques remain grounded in reality for trustworthiness and generalizability; we therefore set out to probe the inner workings of our recently proposed ML model 33 for deriving CCN number concentrations.…”

Reconstructing Aerosols Vertical Profiles with Aggregate Output Learning

“…Collecting high quality observational data of aerosol vertical profiles at large scale is however unfeasible, thus making fully supervised approaches unsuitable. In fact, while in previous work Nair and Yu [9] have addressed the task of CCN number prediction from atmospheric measurements, they resorted to using model data in order to apply fully supervised learning models.…”

Future Directions in Precipitation Science

“…This contrasts with other parameterizations, such as that of convection of the planetary boundary layer (PBL). The need for better or more efficient ways to do computations, through machine learning for instance, has been a constant over recent decades [69][70][71]. One approach has been to use microphysics as a validation tool for physical hypotheses, as it will be difficult to directly include the results of the experiments in the codes.…”