Radiative Effect of Clouds on Tropospheric Chemistry: Sensitivity to Cloud Vertical Distributions and Optical Properties

Clouds play critical roles in influencing tropospheric photochemistry through modification of solar radiation that determines photolysis frequencies. A quantitative understanding of the radiative effect of clouds is required for using global models to assess anthropogenic perturbations to the Earth system.

However, representation of clouds in global models still poses a significant challenge since most cloud processes occur on sub-grid scales and must be parameterized. Uncertainties in cloud distributions and optical properties are therefore a limiting factor in model assessments of the radiative effect of clouds on global tropospheric chemistry.

We have recently assessed the radiative effect of clouds on photolysis frequencies and key oxidants in the troposphere with a global 3-D chemical transport model (GEOS-CHEM) driven by assimilated meteorological observations [Liu et al., 2005]. In this study, we apply the same model to examine the sensitivity of this effect to the uncertainty associated with the distributions and optical properties of clouds.

Hongyu Liu1, James H. Crawford2, Robert B. Pierce2, David B. Considine2, Jennifer A. Logan3, Bryan N. Duncan4,5, Peter Norris4,5, Steven E. Platnick4, Gao Chen2, Robert M. Yantosca3, Mat J. Evans3,6

1National Institute of Aerospace, Hampton, VA; 2NASA Langley Research Center, Hampton, VA; 3Harvard University, Cambridge, MA; 4NASA Goddard Space Flight Center, Greenbelt, MD; 5University of Maryland, Baltimore County, MD; 6Now at University of Leeds, UK

The radiative impact of clouds on global tropospheric chemistry is more sensitive to cloud vertical distribution than to the magnitude of column cloud optical depth.

Model simulations with each of the three (GEOS1-STRAT, GEOS-3, and GEOS-4) cloud distributions all show that tropical upper tropospheric ozone is much less sensitive to the radiative effect of clouds than previously reported by Tie et al. [2003] using the MOZART-2 model (~5% versus ~20-30%).

J[O1D], J[NO2], and OH simulations are very sensitive to cloud absorption of solar radiation. Using 0.99 for cloud single scattering albedo (SSA) in global models would decrease simulated J[O1D], J[NO2], and OH concentrations by ~10-20% in most of the troposphere, relative to SSA=1.0. Realistic values for SSA are between 0.999 and 1.0.

With the launchings of CloudSat and CALIPSO satellites, a unique dataset of cloud optical / physical properties as well as their vertical distributions will be available for improving the constraints on the model calculated radiative effect of clouds on tropospheric chemistry.

3. Evaluation of GEOS-3 Cloud OD with Satellite Observations

Mean COD (grid-scale) for March 2001

MODIS

(MOD08_M3)

ISCCP (D2)

GEOS-3

MODIS

ISCCP

GEOS-3

Probability Distribution Functions

Global Average

GEOS3-OD / MODIS-OD = 0.91

GEOS3-OD / ISCCP-OD = 1.31

4. Intercomparison of GEOS1-S, GEOS-3 and GEOS-4 Cloud ODs

GEOS1-S (1996)

GEOS-3 (2001) GEOS-4 ( 2001)

ISCCPMODIS

GEOS-3

GEOS-4

GEOS1-S

Global Average:

GEOS-3 / GEOS1-S = 5.1; GEOS-3 / GEOS-4 = 1.9

optically

much thinner

changes of global mean OH due to cloudGEOS1-S: -1% GEOS-3: 1% GEOS-4: 14%

GEOS1-S, 1996 GEOS-3, 2001 GEOS-4, 2001

Figure 3. Simulated percentage changes in monthly zonal mean OH concentrations due to the radiative effect of clouds in June. Linear scaling of cloud optical depth with cloud fraction in a grid-box is assumed.

Our model calculations indicate that the changes in global mean OH due to the radiative effect of clouds in June are about -1% (GEOS1-STRAT), 1% (GEOS-3), and 14% (GEOS-4), respectively. The effects on global mean OH are similar for GEOS1-STRAT and GEOS-3 due to similar vertical distributions of clouds, even though the column cloud optical depths in the two archives differ by a factor of about 5. Clouds in GEOS-4 have a much larger impact on global mean OH because more solar radiation is able to penetrate through the optically thin clouds in the upper troposphere, increasing backscattering from low-level clouds. These illustrate that the radiative impact of clouds on global tropospheric OH is more sensitive to cloud vertical distribution than to the magnitude of column cloud optical depth.

Sensitivity of Global Mean OH to COD Magnitude(GEOS-3, 2001)

OH

J[O1D]

SSA=0.99 SSA=0.999

Contact: Hongyu Liu (hyl@nianet.org)

URL: http://research.nianet.org/~hyl

1. Chylek, P., et al., Black carbon and absorption of solar radiation by clouds, J. Geophys. Res., 101(D18), 23,365-23,371, 1996.

2. Hu, Y.X., and K. Stamnes, An accurate parameterization of the radiative properties of water clouds suitable for use in climate models, J. Climate, 6, 728-742, 1993.

3. Liu, H., J.H. Crawford, R.B. Pierce, P. Norris, S.E. Platnick, G. Chen, J.A. Logan, R.M. Yantosca, M.J. Evans, C. Kittaka, Y. Feng, and X. Tie, Radiative effect of clouds on tropospheric chemistry in a global three dimensional chemical transport model, J. Geophys. Res., in revision, Dec. 2005.

4. Park, R.J., D.J. Jacob, B.D. Field, R.M. Yantosca, and M. Chin, Natural and transboundary pollution influences on sulfate-nitrate-ammonium aerosols in the United States: Implications for policy, J. Geophys. Res., 109, D15204, doi:10.1029/2003JD004473, 2004.

5. Tie, X., S. Madronich, S. Walters, R. Zhang, P. Rasch, and W. Collins, Effect of clouds on photolysis and oxidants in the troposphere, J. Geophys. Res., 108(D20), 4642, doi:10.1029/2003JD003659, 2003.

6. Wild, O., X. Zhu, and M.J. Prather, Fast-J: Accurate simulation of in- and below-cloud photolysis in tropospheric chemical models, J. Atmos. Chem., 37, 245-282, 2000.

GEOS-CHEM was driven with a series of meteorological archives (GEOS1-STRAT, GEOS-3, and GEOS-4) generated by the Goddard Earth Observing System data assimilation system (GEOS DAS) at the NASA Global Modeling and Assimilation Office (GMAO), which have significantly different cloud optical depths (OD) and vertical distributions.

Ozone-NOx-CO-VOC coupled to aerosol chemistry [Park et al., 2004].

Photolysis rate calculation: Fast-J radiative transfer algorithm [Wild et al., 2000] with 3-D cloud optical depth and cloud fraction taken from GEOS.

Model extensively evaluated with surface, in-situ, and satellite observations (http://www-as.harvard.edu/chemistry/trop/geos).

Sensitivity simulations: 1996 (GEOS1-STRAT); August 2000 – December 2001 (GEOS-3 and GEOS-4); 4ox5o horizontal resolution. The radiative effect of clouds is represented by subtraction of the clear-sky simulation from the cloudy-sky simulation.

Figure 1. Left panels: The global distribution of GEOS-3 monthly mean (grid-scale, column) cloud optical depths compared to MODIS and ISCCP retrievals for March 2001 [Liu et al., 2005]. Right panel: Same as left panel, but shown as probability distribution functions. GEOS-3 cloud optical depths show peaks in the tropics associated with deep convective clouds and at midlatitudes associated with extratropical cyclones in NH and marine stratiform clouds in SH. These features reasonably agree with MODIS and ISCCP cloud retrieval products, although GEOS-3 tends to overestimate cloud ODs in the tropics and SH midlatitudes [Liu et al., 2005]. Also see Figure 2.

Figure 2. Top left panel: Intercomparison of the June monthly zonal mean column cloud optical depths in GEOS1-STRAT (1996), GEOS-3 (2001) and GEOS-4 (2001). MODIS and ISCCP retrievals are for June 2001. Other panels (color images): Latitude-height cross sections of cloud optical depths per kilometer for June.

The column cloud optical depths in GEOS-3 generally agree with the satellite retrieval products from MODIS and ISCCP within +/-5-30%, while those in GEOS1-STRAT and GEOS-4 are too low by factors of about 5 and 2, respectively. With respect to vertical distribution, clouds in GEOS-4 are optically much thinner in the tropical upper troposphere compared to those in GEOS1-STRAT and GEOS-3.

5. Sensitivity to Three Cloud Vertical Distributions and OD

6. Sensitivity to Progressively Adjusted Cloud Optical Depth

Figure 5. Simulated percentage changes in global mean OH due to the radiative effect of clouds as the magnitude of 3-D cloud optical depths is progressively adjusted. A cloud OD factor of 0.5 corresponds to half of the original GEOS-3 cloud optical depth with the same 3-D spatial distributions. Global average effects are modest for all cloud OD factors, reflecting the opposite effects of enhanced (weakened) photochemistry above (below) clouds. Monotonic increases for January/October reflect the dominant backscattering from low-level clouds. Non-monotonic changes for March/June are a result of the sufficiently large optical depths due to high clouds which allow less solar radiation to penetrate down to the lower levels and thus limit backscattering from low-level clouds. See Figure 6 for latitude-height cross sections of GEOS-3 cloud ODs in different seasons.

Figure 6. Latitude-height cross sections of GEOS-3 monthly mean cloud optical depths for different seasons.

7. Sensitivity to Cloud Absorption of Solar Radiation

Figure 7. Sensitivity of simulated J[O1D] and OH to cloud absorption of solar radiation for June (GEOS-3). The plots show the percentage changes in monthly zonal mean J[O1D] and OH concentrations when cloud SSA=0.99 (left panels) and SSA=0.999 (right panels), relative to SSA=1.0. Results indicate that 1% decrease in SSA would decrease J[O1D] and OH concentrations by ~10-20% in most of the troposphere.

Figure 4. Same as Figure 3, but for ozone (June). Contour levels are –5, -2, -1, 0, 1, 2, 5, 10%. Tropical upper tropospheric ozone is much less sensitive to the radiative effect of clouds than previously reported by Tie et al. [2003] using the MOZART-2 model (~5% versus ~20-30%).

A cautionary note is presented here that 0.99 is too low a value for cloud single scattering albedo (SSA) and is not consistent with current knowledge of cloud absorption of solar radiation in the UV wavelength relevant to tropospheric chemistry. SSA for pure water droplets is between 0.999990 and 0.999999 (i.e., little absorption) in the UV wavelength range [Hu and Stamnes, 1993]. SSA for contaminated clouds containing black carbon is still between 0.999 and 0.9999 at 550nm [Chylek et al., 1996]. SSA=0.99 however was used in some recent literature of tropospheric chemistry.

GEOS1-S, 1996 GEOS-3, 2001 GEOS-4, 20011. Background and Objective

2. Model and Method

9. References

8. Summary and Conclusions

Paper Number:

A51C-0062