Linking large-scale circulation patterns to low-cloud properties

Abstract

Abstract. The North Pacific High (NPH) is a fundamental meteorological feature present during the boreal warm season. Marine boundary layer (MBL) clouds, which are persistent in this oceanic region, are influenced directly by the NPH. In this study, we combine 11 years of reanalysis and an unsupervised machine learning technique to examine the gamut of 850 hPa synoptic-scale circulation patterns. This approach reveals two distinguishable regimes – a dominant NPH setup and a land-falling cyclone – and in between a spectrum of large-scale patterns. We then use satellite retrievals to elucidate for the first time the explicit dependence of MBL cloud properties (namely cloud droplet number concentration, liquid water path, and shortwave cloud radiative effect – CRESW) on 850 hPa circulation patterns over the northeast Pacific Ocean. We find that CRESW spans from −146.8 to −115.5 W m−2, indicating that the range of observed MBL cloud properties must be accounted for in global and regional climate models. Our results demonstrate the value of combining reanalysis and satellite retrievals to help clarify the relationship between synoptic-scale dynamics and cloud physics.

Linking Synoptic Patterns to Cloud Properties and Local Circulations Over Southeastern Texas

“…In terms of the SOM initial parameters, additional tests are conducted by decreasing the number of iterations from 5,000 to 2,500, which turns out to have a relatively small impact on the SOM performance and the identified synoptic patterns. These results are consistent with previous studies (e.g., Cassano et al, 2006;Johnson et al, 2008;Juliano & Lebo, 2020;Skific et al, 2009).…”

“…Such methods have shown success in capturing the variation of large‐scale dynamics, especially during the transitional weather states. This concept has been successfully applied to several DOE ARM sites such as the Southern Great Plains site and the Eastern North Atlantic site to study large‐scale controls on cloud properties (e.g., Ford et al., 2015; Juliano & Lebo, 2020; Mechem et al., 2018; Song et al., 2019). The SOM is accomplished using an open‐source Python package, minisom (Vettigli, 2018), which has been widely used in the science community (e.g., Gorgoglione et al., 2021; Lessin et al., 2020).…”

“…The initial learning rate is set at 0.5 and the maximum number of iterations is 5,000 following Song et al. (2019) and Juliano and Lebo (2020). The initial σ is set to 1.…”

“…The lr(t) and σ(t) are in the form of a decreasing function, lr(t) = lr/(1 + t/T), σ(t) = σ/ (1 + t/T), where T is half of the maximum number of iterations. The initial learning rate is set at 0.5 and the maximum number of iterations is 5,000 following Song et al (2019) and Juliano and Lebo (2020). The initial σ is set to 1.…”

Clustering of large‐scale circulation patterns during tropical cyclone preformation period using self‐organizing maps

“…The SOM (Kohonen, 1982) is a neural network-based, unsupervised machine learning technique, especially suitable for the analysis of high-dimensional statistical data. Given a two-dimensional input array of multiple sample vectors and an output topology structure (e.g., Juliano and Lebo, 2020), the SOM algorithm can automatically organize a topology through iteratively mapping similar input vectors close to each other.…”