Quantifying radiative effects of the cloud–clear-sky transition using ground-based infrared imagery

Clouds exert a fundamental control over the Earth’s radiative balance by modulating both incoming shortwave solar radiation and outgoing longwave terrestrial radiation, resulting in a net radiative cooling of approximately 19 Wm^-2. In contrast to well-mixed greenhouse gases such as CO₂ and methane, the radiative impact of clouds exhibits strong spatial and temporal variability, making it intrinsically difficult to quantify and one of the dominant sources of uncertainty in contemporary climate models.

This study addresses a complementary and often overlooked aspect of this problem: the cloud–clear-sky interface, characterized by a continuous transition between cloudy and cloud-free conditions. Due to this gradual transition, defining the boundaries of a cloud region is not straightforward and depends strongly on the observational context. In numerical models, clouds are typically defined using relative humidity thresholds, whereas satellite-based cloud detection relies on radiance thresholds that vary across spectral bands.

Here, we analyze ground-based thermal infrared imagery (10–12 µm) by comparing observations with modeled clear-sky radiances. Using radiance exceedances relative to clear-sky emission, we quantify the radiative effect associated with the cloud–clear-sky transition. A preliminary analysis based on a limited subset of the available dataset indicates that the selection of cloud spectral radiance thresholds can lead to differences of approximately 0.4 Wm^-2µm^-1 when compared with definitions that classify clouds only under high-confidence conditions. A comprehensive analysis will be completed prior to the conference.

The data were collected at four distinct sites: the Atmospheric Radiation Measurement (ARM) Southern Great Plains (SGP) observatory in Oklahoma, USA; the Federal University of São Paulo (UNIFESP), Diadema campus, located in the metropolitan region of São Paulo, Brazil; and two sites in the Amazon region—Amazonas State University (UEA), Escola Normal Superior, near downtown Manaus, Brazil, and Embrapa Amazônia Ocidental, situated in a rural area of Manaus. Measurements conducted in the Amazon region are part of the project “Measurements of cloud properties relevant to improving the prediction of intense rainfall in Manaus”, funded by Fundação de Amparo à Pesquisa do Estado do Amazonas (FAPEAM).