[1] Planetary boundary layer (PBL) processes control energy, water, and pollutant exchanges between the surface and free atmosphere. However, there is no observationbased global PBL climatology for evaluation of climate, weather, and air quality models or for characterizing PBL variability on large space and time scales. As groundwork for such a climatology, we compute PBL height by seven methods, using temperature, potential temperature, virtual potential temperature, relative humidity, specific humidity, and refractivity profiles from a 10 year, 505-station radiosonde data set. Six methods are directly compared; they generally yield PBL height estimates that differ by several hundred meters. Relative humidity and potential temperature gradient methods consistently give higher PBL heights, whereas the parcel (or mixing height) method yields significantly lower heights that show larger and more consistent diurnal and seasonal variations (with lower nighttime and wintertime PBLs). Seasonal and diurnal patterns are sometimes associated with local climatological phenomena, such as nighttime radiation inversions, the trade inversion, and tropical convection and associated cloudiness. Surface-based temperature inversions are a distinct type of PBL that is more common at night and in the morning than during midday and afternoon, in polar regions than in the tropics, and in winter than other seasons. PBL height estimates are sensitive to the vertical resolution of radiosonde data; standard sounding data yield higher PBL heights than high-resolution data. Several sources of both parametric and structural uncertainty in climatological PBL height values are estimated statistically; each can introduce uncertainties of a few 100 m.
Recently, state-of-the-art classification performance of natural images has been obtained by self-supervised learning (S2L) as it can generate latent features through learning between different views of the same images. However, the latent semantic information of similar images has hardly been exploited by these S2L-based methods. Consequently, to explore the potential of S2L between similar samples in hyperspectral image classification (HSIC), we propose the nearest neighboring self-supervised learning (N2SSL) method, by interacting between different augmentations of reliable nearest neighboring pairs (RN2Ps) of HSI samples in the framework of bootstrap your own latent (BYOL). Specifically, there are four main steps: pretraining of spectral spatial residual network (SSRN)-based BYOL, generation of nearest neighboring pairs (N2Ps), training of BYOL based on RN2P, final classification. Experimental results of three benchmark HSIs validated that S2L on similar samples can facilitate subsequent classification. Moreover, we found that BYOL trained on an un-related HSI can be fine-tuned for classification of other HSIs with less computational cost and higher accuracy than training from scratch. Beyond the methodology, we present a comprehensive review of HSI-related data augmentation (DA), which is meaningful to future research of S2L on HSIs.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.