Understanding the dependence of tropical high cloud amount and radiative flux on sea surface temperatures

Abstract

We use observations and four GFDL AGCMs to analyze the relation between variations in spatial patterns and area-averaged quantities in the top-of-atmosphere radiative fluxes, cloud amount and precipitation related to El Nino over the period 1979-2008. We find that model-to-observation biases in the base state translate into corresponding biases in anomalies in response to El Nino. The pattern and amplitude of model biases in reflected shortwave (SW) and outgoing longwave radiation (OLR) follows expectations based on their biases in cloud amount: models with a positive cloud amount bias, compared to observations, have too strong local responses to El Nino in cloud amount, SW, OLR and precipitation. Tropical average OLR increases in response to El Nino in observations and models (correlation coefficients (r) with Nino 3.4 Index in range 0.4 to 0.6). Weaker correlations are found for SW (r: -0.6 to 0), cloud amount (r: -0.2 to +0.1) and precipitation (r: -0.2 to 0).

Tropical high clouds are closely coupled to deep convection, but local cloud amount and convective mass flux are non-linearly related. We use the GFDL-AM2 model forced with idealized SST perturbations to study the sensitivity of high clouds to the large-scale distribution of convection.

Increasing/decreasing the SST contrast between convective and

non-convective regions decreases/increases the tropical deep convective

area, and warming of convective areas decreases the tropical average convective mass flux (<m_c>). In all experiments, fractional high cloud amount changes are less than fractional changes in <m_c>. High cloud amount is half as sensitive as expected from the climatological average cloud amount, as a function of convective mass flux, due to strong compensation from non-convective high clouds. The latter results from changes in relative humidity related to the change in <m_c>. This effect renders high cloud amount remarkably robust to perturbations, though radiative effects of convective and non-convective clouds will differ.

Finally, we analyze tropospheric temperature trends in 19 atmosphere-only models and show the spread is due, at least in part, to difference in precipitation-weighted SST (P.SST) trends (r=+0.99). We confirm vertical amplification of surface warming by a factor of 2.4 at the 300mb level, relative to P.SST.