Estimation of Cloud Condensation Nuclei (CCN) from SPEXone on PACE using a neural network retrieval algorithm: Comparison to AERONET and ATLID/EarthCARE

Cloud condensation nuclei (CCN) play a critical role in aerosol–cloud interactions (ACI). It has been shown that the column number of aerosol particles exceeding a predetermined threshold radius (NCCN) is a suitable CCN proxy. Previously this CCN proxy has been estimated from PARASOL using Level-2 aerosol microphysical and/or optical property retrievals. With the launch of SPEXone on Plankton, Aerosol, Cloud, ocean Ecosystem (PACE) mission, further improvements in NCCN retrievals can be achieved. In particular, retrieved refractive index can enable estimation of the volume fraction of aerosol-water, facilitating the derivation of dry aerosol size distibution and correspondingly dry CCN. In addition, retrieved aerosol layer height (ALH) can be used to estimate the boundary layer (BL) contribution of NCCN (NCCN(BL)).We developed a deep neural network (NN) algorithm as an extension of the Remote sensing of Trace gas and Aerosol Products (RemoTAP)-NN framework to directly retrieve dry NCCN, and NCCN(BL) from SPEXone measurements. The algorithm was trained on synthetic SPEXone measurements generated from a three-mode aerosol representation including fine mode, insoluble coarse/dust mode, and soluble coarse mode. Initial validation was performed using independent synthetic measurements, based on the ECHAM-HAM global aerosol-climate model.For validating NCCN from real SPEXone observations, we use collocated AERONET data, for which both dry and ambient NCCN are computed. On the log base 10 scale, the NN algorithm achieved RMSDs of 0.33 (dry) and 0.21 (ambient) over land, and 0.21 (dry) and 0.20 (ambient) over ocean. The slightly higher RMSD for dry NCCN is attributed to the cases in which the AERONET derived refractive index reaches its upper limit of 1.6. In comparison, CCN proxies derived using the classical RemoTAP algorithm exhibited RMSDs approximately 20% higher. Ongoing work focuses on validating the retrieved fraction of aerosols within the boundary layer using EarthCARE ATLID Level-2 observations.