Supplementary material for Investigating relationshipsbetween aerosol optical depth and cloud fraction usingsatellite, aerosol reanalysis and general circulation

model data

B. S. Grandey, P. Stier and T. M. Wagner

February 8, 2013

(a) MODIS–MACC t+3 Sampled (b) MODIS–MACC t+24 Sam-pled

(c) Tompkins Control

(d) Tompkins NoAIE (e) Tompkins Control dry (f) Tompkins NoConvScav

Figure S1: Same as Figure 2 but for (a), (b) MODIS–MACC sampled according to MODIS τavailability and (c)–(f) ECHAM5-HAM Tompkins simulations.

1

Sundqvis

t

Ocean: AWM= 0.15, Med= 0.13 Land: AWM= 0.18, Med= 0.14

(a) Control τ

Ocean: AWM= 0.04, Med= 0.03 Land: AWM= 0.09, Med= 0.06

(b) Control dry τ

Ocean: AWM= 0.39, Med= 0.29 Land: AWM= 0.42, Med= 0.33

(c) NoConvScav τ

0.10 0.20 0.30 0.40 0.50 0.60

Figure S2: Annual (all seasons) mean aerosol optical depth (τ) for ECHAM5-HAM Sundqvistsimulations. The area-weighted means (AWM) and median (Med) for both ocean and land areshown beneath each map. The Tompkins simulations have very similar annual mean τ fields tothe respective Sundqvist simulations. The NoAIE simulations have very similar annual meanτ fields to the Control simulations. The area-weighted mean (AWM) and median (Med) forboth ocean and land are shown beneath each map.

2

Ocean: EWM= 0.179, AWM= 0.191, Med= 0.188Land: EWM= 0.137, AWM= 0.147, Med= 0.143

(a) Aqua-MODIS Coll. 5

Ocean: EWM= 0.098, AWM= 0.188, Med= 0.179Land: EWM= 0.125, AWM= 0.177, Med= 0.158

(b) MODIS–MACC t+3hour

Ocean: EWM= 0.065, AWM= 0.139, Med= 0.112Land: EWM= 0.081, AWM= 0.139, Med= 0.120

(c) MODIS–MACC t+24 hour

Ocean: EWM= −0.004, AWM= 0.004, Med=−0.009Land: EWM= −0.003, AWM= −0.007, Med=−0.004

(d) Sundqvist Control

Ocean: EWM= −0.004, AWM= 0.002, Med=−0.008Land: EWM= 0.001, AWM= −0.003, Med= 0.001

(e) Sundqvist NoAIE

Ocean: EWM= −0.062, AWM= −0.064, Med=−0.057Land: EWM= −0.021, AWM= −0.023, Med=−0.018

(f) Sundqvist Control Dry

Ocean: EWM= 0.005, AWM= 0.070, Med= 0.025Land: EWM= 0.004, AWM= 0.008, Med= 0.007

(g) Sundqvist NoConvScav

−0.400−0.300−0.200−0.100 0.000 0.100 0.200 0.300 0.400

Figure S3: Same as Figures 1 and 2, but for dfcd ln τ

. The grid-method of Grandey and Stier (2010)has been used. The lin–log relationship was chosen based on semi-empirical considerations(Chapter 3 of Grandey, 2011). White regions are where the data are not significantly differentfrom zero at two-sigma confidence. The error-weighted mean (EWM), area-weighted mean(AWM) and median (Med) for both ocean and land are shown beneath each map.

3

Ocean: AWM= 0.43, Med= 0.43 Land: AWM= 0.41, Med= 0.43

(a) Aqua-MODIS Coll. 5

Ocean: AWM= 0.27, Med= 0.26 Land: AWM= 0.27, Med= 0.24

(b) MODIS–MACC t+3hour

Ocean: AWM= 0.19, Med= 0.18 Land: AWM= 0.20, Med= 0.19

(c) MODIS–MACC t+24 hour

Ocean: AWM=−0.03, Med=−0.03 Land: AWM=−0.01, Med= 0.00

(d) Sundqvist Control

Ocean: AWM=−0.04, Med=−0.04 Land: AWM=−0.01, Med= 0.00

(e) Sundqvist NoAIE

Ocean: AWM=−0.13, Med=−0.13 Land: AWM=−0.05, Med=−0.04

(f) Sundqvist Control Dry

Ocean: AWM= 0.08, Med= 0.06 Land: AWM= 0.02, Med= 0.03

(g) Sundqvist NoConvScav

−0.800−0.600−0.400−0.200 0.000 0.200 0.400 0.600 0.800

Figure S4: Same as Figures 1 and 2, but for dfc–d ln τ correlation.

4

References

Grandey, B. S.: Investigating aerosol–cloud interactions, Ph.D. thesis,University of Oxford, UK, URL http://ora.ox.ac.uk/objects/uuid:

8b48c02b-3d43-4b04-ae55-d9885960103d, 2011.

Grandey, B. S. and Stier, P.: A critical look at spatial scale choices in satellite-based aerosol in-direct effect studies, Atmos. Chem. Phys., 10, 11 459–11 470, doi:10.5194/acp-10-11459-2010,2010.

5