Premium
Sensitivity of one‐dimensional radiative biases to vertical cloud‐structure assumptions: Validation with aircraft data
Author(s) -
Di Giuseppe F.
Publication year - 2005
Publication title -
quarterly journal of the royal meteorological society
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.744
H-Index - 143
eISSN - 1477-870X
pISSN - 0035-9009
DOI - 10.1256/qj.03.129
Subject(s) - radiative transfer , liquid water content , radiative flux , adiabatic process , computational physics , monte carlo method , parametrization (atmospheric modeling) , atmospheric radiative transfer codes , physics , atmospheric sciences , environmental science , mechanics , mathematics , optics , cloud computing , thermodynamics , statistics , computer science , operating system
Three representations of an observed stratocumulus system are generated by combining aircraft observations with a simple statistical model. The realizations differ in their representation of the vertical cloud structure while the horizontal variability is identical. In the control case (A) both the adiabatic liquid‐water profile and the effect of wind‐shear induced vertical decorrelation are represented. The second simulation (B) removes the wind‐shear effect by assuming maximum overlap between adjacent layers. The third case (C) instead removes vertical variability by averaging the in‐cloud liquid water for each column. For each of these scenes Monte Carlo simulated solar fluxes are compared against observed flux measurements. Cases A and B agree with observed (horizontal) flux variations within statistical uncertainty, while case C, which neglects vertical variability, is not able to reproduce the observed fluxes. The comparison between the radiative fields produced by these three representations of the stratocumulus system, calculated using a three‐dimensional radiative‐transfer solution, an independent pixel approximation (IPA) and a plane‐parallel (PP) approach, shows substantial differences. Not accounting for the adiabatic liquid‐water profile generates a systematic increase in the optical depth, τ when the effective radius is computed from mean liquid‐water content and droplet‐number concentration, that can be responsible for a 5% increase in the reflection for shallow boundary‐layer cloud systems (τ≈1). A much stronger effect in the radiative properties is produced by varying the cloud‐overlap rule applied. While changing from maximum to random overlap does not introduce any variation in the optical depth of the cloud scene, it does introduce an increase in the reflection that is proportional to the relative change in total cloud fraction. The magnitude of these latter biases is comparable to that produced by unresolved horizontal variability. Moreover, it is shown that, when the vertical cloud structure is properly resolved, the effect of horizontal fluctuations is also reduced. Copyright © 2005 Royal Meteorological Society