Export 24 results:
Sort by: Author Title [ Type  (Asc)] Year
Book Chapter
Somerville, RCJ, Iacobellis SF.  1987.  Cloud-radiation interactions: Effects of cirrus optical thickness feedbacks. Short- and Medium-Range Numerical Weather Prediction. ( Matsuno T, Ed.).:177-185., [Tokyo]: Meteorological Society of Japan Abstract
Lane, DE, Somerville RCJ, Iacobellis S.  2001.  Evaluation of a Stochastic Radiative Transfer Model Using Ground-based Measurements. IRS 2000: Current Problems in Atmospheric Radiation : Proceedings of the International Radiation Symposium, St. Petersberg, Russia, 24-29 July 2000. ( Smith WL, Timofeyev YM, Eds.).:245-248.: A Deepak Publishing Abstract
Iacobellis, S, Somerville RCJ, Lane DE.  2001.  SCM Sensitivity to Microphysics, Radiation, and Convection Algorithms. IRS 2000: Current Problems in Atmospheric Radiation : Proceedings of the International Radiation Symposium, St. Petersberg, Russia, 24-29 July 2000. ( Smith WL, Timofeyev YM, Eds.).:1287-1290.: A Deepak Publishing Abstract
IPCC.  2007.  Summary for Policymakers. Climate change 2007 : the physical science basis : contribution of Working Group I to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change. ( Solomon S, Qin D, Manning M, Chen Z, Marquis M, Averyt K, Tignor M, Miller H, Eds.)., Cambridge; New York: Cambridge University Press Abstract
Conference Paper
Chertock, B, Iacobellis S, Somerville C.  1987.  Remote sensing studies of oceanic cloud-radiation feedbacks. Atmospheric radiation progress and prospects (Proceedings of the Beijing International Radiation Symposium, August 26-30, 1986). ( Liou K, Chou H, Eds.).:508-514., Beijing, China: Science Press and American Meteorological Society Abstract
Journal Article
Isakari, SM, Somerville RCJ.  1989.  Accurate numerical solutions for Daisyworld. Tellus Series B-Chemical and Physical Meteorology. 41:478-482.   10.1111/j.1600-0889.1989.tb00324.x   AbstractWebsite

The numerical solutions of the Daisyworld model of Watson and Lovelock contain significant quantitative errors. We give accurate numerical solutions for the same cases. We also show how the errors may have been caused by failure to enforce computational constraints such as strict tests of steadiness. The errors which we find do not qualitatively alter the main conclusions of Watson and Lovelock, but they illustrate a peril. The Daisyworld model is an example of a mathematical system which is too idealized to be compared with observations but too complex to be solved analytically. Such systems can be probed only by numerical simulations, so it is crucial that the computations be trustworthy.

Lee, WH, Iacobellis SF, Somerville RCJ.  1997.  Cloud radiation forcings and feedbacks: General circulation model tests and observational validation. Journal of Climate. 10:2479-2496.   10.1175/1520-0442(1997)010<2479:crfafg>;2   AbstractWebsite

Using an atmospheric general circulation model (the National Center for Atmospheric Research Community Climate Model: CCM2), the effects on climate sensitivity of several different cloud radiation parameterizations have been investigated. In addition to the original cloud radiation scheme of CCM2, four parameterizations incorporating prognostic cloud water were tested: one version with prescribed cloud radiative properties and three other versions with interactive cloud radiative properties. The authors' numerical experiments employ perpetual July integrations driven by globally constant sea surface temperature forcings of two degrees, both positive and negative. A diagnostic radiation calculation has been applied to investigate the partial contributions of high, middle, and low cloud to the total cloud radiative forcing, as well as the contributions of water vapor, temperature, and cloud to the net climate feedback. The high cloud net radiative forcing is positive, and the middle and low cloud net radiative forcings are negative. The total net cloud forcing is negative in all of the model versions. The effect of interactive cloud radiative properties on global climate sensitivity is significant. The net cloud radiative feedbacks consist of quite different shortwave and longwave components between the schemes with interactive cloud radiative properties and the schemes with specified properties. The increase in cloud water content in the warmer climate leads to optically thicker middle-and low-level clouds and in turn to negative shortwave feedbacks for the interactive radiative schemes, while the decrease in cloud amount simply produces a positive shortwave feedback for the schemes with a specified cloud water path. For the longwave feedbacks, the decrease in high effective cloudiness for the schemes without interactive radiative properties leads to a negative feedback, while for the other cases, the longwave feedback is positive. These cloud radiation parameterizations are empirically validated by using a single-column diagnostic model, together with measurements from the Atmospheric Radiation Measurement program and from the Tropical Ocean Global Atmosphere Combined Ocean-Atmosphere Response Experiment. The inclusion of prognostic cloud water produces a notable improvement in the realism of the parameterizations, as judged by these observations. Furthermore, the observational evidence suggests that deriving cloud radiative properties from cloud water content and microphysical characteristics is a promising route to further improvement.

Ghan, S, Randall D, Xu KM, Cederwall R, Cripe D, Hack J, Iacobellis S, Klein S, Krueger S, Lohmann U, Pedretti J, Robock A, Rotstayn L, Somerville R, Stenchikov G, Sud Y, Walker G, Xie SC, Yio J, Zhang MH.  2000.  A comparison of single column model simulations of summertime midlatitude continental convection. Journal of Geophysical Research-Atmospheres. 105:2091-2124.   Doi 10.1029/1999jd900971   AbstractWebsite

Eleven different single-column models (SCMs) and one cloud ensemble model (CEM) are driven by boundary conditions observed at the Atmospheric Radiation Measurement (ARM) program southern Great Plains site for a 17 day period during the summer of 1995. Comparison of the model simulations reveals common signatures identifiable as products of errors in the boundary conditions. Intermodel differences in the simulated temperature, humidity, cloud, precipitation, and radiative fluxes reflect differences in model resolution or physical parameterizations, although sensitive dependence on initial conditions can also contribute to intermodel differences. All models perform well at times but poorly at others. Although none of the SCM simulations stands out as superior to the others, the simulation by the CEM is in several respects in better agreement with the observations than the simulations by the SCMs. Nudging of the simulated temperature and humidity toward observations generally improves the simulated cloud and radiation fields as well as the simulated temperature and humidity but degrades the precipitation simulation for models with large temperature and humidity biases without nudging. Although some of the intermodel differences have not been explained, others have been identified as model problems that can be or have been corrected as a result of the comparison.

Iacobellis, SF, Somerville RCJ.  1991.  Diagnostic modeling of the Indian monsoon onset: Part 1: Model description and validation. Journal of the Atmospheric Sciences. 48:1948-1959.   10.1175/1520-0469(1991)048<1948:dmotim>;2   AbstractWebsite

A new type of diagnostic model is developed and applied to the study of the onset of the Indian summer monsoon. The purpose of the model is to aid in the analysis of interactions between the physical processes that affect the monsoon onset. The model is one-dimensional and consists of a single atmospheric column coupled to an ocean mixed layer. The atmospheric component of the model includes representations of all the physical processes typically included in general circulation models, except that the fields of vertical motion and horizontal advection are specified at each time step from observational data rather than predicted. With these time-dependent observational inputs, the model is then integrated numerically to produce consistent profiles of atmospheric temperature and humidity, together with energy budget components and other diagnostic quantities. The atmospheric model is based on the thermodynamic energy equation and a conservation equation for water. Parameterizations of the effects of solar and terrestrial radiation, interactive cloudiness, convection, condensation, surface fluxes, and other processes are adapted from current practice in numerical weather prediction and general circulation modeling. The model includes 15 layers in the vertical and employs a time step of 1 hour. Results are presented from four-week integrations at different locations over the Arabian Sea during the 1979 monsoon onset period. Comparison of model results with independent observational data shows that the model demonstrates considerable skill in reproducing the large increase in precipitation associated with the monsoon onset, together with significant changes in surface fluxes, cloudiness, and other variables. This realism suggests that the model is a promising tool for achieving an increased understanding of the role of interacting physical processes and for developing improved prognostic models for simulating the monsoon onset.

Iacobellis, SF, Somerville RCJ.  1991.  Diagnostic modeling of the Indian monsoon onset: Part 2: Budget and sensitivity studies. Journal of the Atmospheric Sciences. 48:1960-1971.   10.1175/1520-0469(1991)048<1960:dmotim>;2   AbstractWebsite

A one-dimensional diagnostic coupled air-sea model (described in the companion paper) is applied to the analysis of the heat and moisture budgets over the Arabian Sea during the 1979 monsoon onset period. The surface energy budget, which is dominated by a balance between net shortwave radiation and latent heat during the preonset period, is significantly altered just prior to the onset itself. At that time, cloud cover sharply increases and the net shortwave flux correspondingly decreases. Subsequently, increasing surface winds produce a large increase in the latent heat flux a few days after the onset. In the free atmosphere, the heat budget displays a similarly dramatic change. At 500 mb, radiative fluxes and horizontal and vertical advection dominate the heat budget before the onset. After the onset, however, the budget is primarily a balance between deep convective heating and vertical advective cooling. The 500-mb moisture budget displays a correspondingly strong effect. Before the onset, horizontal advection of moisture is the dominant term, while after the onset, the distribution by convection of the surface moisture flux, together with moisture removal by large-scale condensation, becomes important. Sensitivity studies with the model illuminate the role of interacting physical processes. Model results show that the moistening due to horizontal advection tends to alter the radiative fluxes so as to hinder the formation and maintenance of the inversion that characterizes preonset conditions, thus favoring the formation of deep convection. This result is consistent with a suggestion by Doherty and Newell. Additionally, the interaction between the atmosphere and the upper ocean is explored in a series of sensitivity experiments. The decrease in ocean mixed-layer temperature, which follows the monsoon onset, acts to reduce the latent heat flux significantly. This effect may influence the duration and intensity of the monsoon, as well as the total precipitation, and underscores the potential importance of an accurate specification of sea surface temperature for monsoon prediction.

Iacobellis, SF, Frouin R, Somerville RCJ.  1999.  Direct climate forcing by biomass-burning aerosols: Impact of correlations between controlling variables. Journal of Geophysical Research-Atmospheres. 104:12031-12045.   10.1029/1999jd900001   AbstractWebsite

Estimates of the direct climate forcing by condensed organic species resulting from biomass burning have been made using bulk radiative transfer models of various complexity and the SUNRAY radiation code of the European Centre for Medium-Range Weather Forecasts general circulation model. Aerosols arising from the burning of tropical forests and savannas as well as those from biomass fires outside the tropics are considered. The bulk models give values ranging from -1.0 to -0.6 W m(-2), which compare with -0.7 W m(-2) using the SUNRAY code. There appears to be significant uncertainty in these values due to uncertainties in the model input parameters. The difference is only 13% between the forcing obtained by taking into account the spatial and temporal distribution of the controlling variables and the forcing obtained using global averages fur all the variables. This indicates that the effects of variations in the controlling variables tend to compensate. Yet the forcing varies by up to 34% depending on which variables are set to global averages. The SUNRAY results show that the efficiency at which the biomass-burning aerosols backscatter sunlight in cloudy conditions is 0.53, a value significantly higher than that reported for sulfate aerosols. Most of the difference is due to the relatively low latitude (hence low sun zenith angle) of the biomass-burning aerosol sources relative to the sulfate aerosol sources. The implication is that clouds should not be assumed to have a reflectivity of unity in bulk models. Comparison of SUNRAY and bulk model results points to other potential problems with bulk models. First, the use in bulk models of mean aerosol optical properties across the entire solar spectrum has significant impact on the calculated forcing and may account for 23% of the difference between SUNRAY and bulk model estimates in clear-sky conditions. Second, neglecting multiple scattering in bulk models introduces significant differences in the clear-sky forcing at high sun zenith angles.

Somerville, RCJ, Iacobellis S, Lee WH.  1996.  Effects of cloud-radiation schemes on climate model results. World Resource Review. 8:321-333. Abstract

A current dilemma of climate modeling is that model results are strongly sensitive to the treatment of certain poorly-understood physical processes, especially cloud-radiation interactions. Thus, different models with alternative plausible parameterizations often give widely varying results. Yet, we typically have had little basis for estimating which parameterization is more realistic. Of the many physical processes involved in climate simulations, feedbacks due to cloud-radiation interactions are thought to be the largest single source of uncertainty. In fact, most of the global differences in results between leading climate models, as measured by their sensitivity to greenhouse gases, can be traced to different model treatments of cloud-radiation interactions.Using a modern atmospheric general circulation model (the National Center for Atmospheric Research Community Climate Model: CCM2), we have investigated the effects on climate sensitivity of several different cloud-radiation parameterizations. At the same time, we have validated these parameterizations directly with observations from field experiments. In addition to the original cloud-radiation scheme of CCM2, we tested four parameterizations incorporating prognostic cloud water: one version with prescribed cloud radiative properties and three other versions with interactive cloud radiative properties. Comparisons with measurements suggest that schemes with explicit cloud water budgets and interactive radiative properties are potentially capable of matching observational data closely.

Iacobellis, SF, Somerville RCJ.  2006.  Evaluating parameterizations of the autoconversion process using a single-column model and Atmospheric Radiation Measurement Program measurements. Journal of Geophysical Research-Atmospheres. 111   10.1029/2005jd006296   AbstractWebsite

A single-column model is used to evaluate the performance of two types of autoconversion parameterizations. The model results are compared to data collected at the Atmospheric Radiation Measurement Program's Southern U. S. Great Plains site. The model is run over a period covering 2 years (2000-2001), and the results are analyzed for time periods varying from hourly to seasonal. During a relatively short 27-hour period during March 2000 characterized primarily by shallow frontal clouds, modeled values of cloud liquid water were better simulated using a Manton-Cotton-type autoconversion parameterization. However, over longer timescales representing a multitude of different cloud types and meteorological conditions, a Sundqvist-type parameterization produced better results. Analysis of the model results indicates that the Manton-Cotton-type parameterization does better during periods when shallow clouds are present without any overlying clouds, while the Sundqvist-type parameterization is preferred during periods when high and low clouds coexist. A possible explanation is that precipitation from high clouds may not be represented well by the SCM, thus affecting the precipitation formation rates in any lower clouds. Sensitivity tests using the Manton-Cotton parameterization indicate that the autoconversion rate is sensitive to the specification of the cloud droplet number concentration (N-c). The single-column model, as well as many general circulation models, specify N-c as a constant value. However, limited in situ measurements suggest that N-c varies significantly in time. The mean modeled top-of-atmosphere cloud radiative forcing during the 2-year period 2000-2001 differed by 3 W m(-2) as the cloud droplet concentration was varied between minimum and maximum values suggested by the in situ measurements. These results imply that model-produced hydrological cycle and cloud-radiation interactions could be better modeled using an accurate time-dependent measure of the cloud droplet concentration.

Iacobellis, SF, Somerville RCJ.  2000.  Implications of microphysics for cloud-radiation parameterizations: Lessons from TOGA COARE. Journal of the Atmospheric Sciences. 57:161-183.   10.1175/1520-0469(2000)057<0161:iomfcr>;2   AbstractWebsite

A single-column model (SCM) and observational data collected during TOGA COARE were used to investigate the sensitivity of model-produced cloud properties and radiative fluxes to the representation of cloud microphysics in the cloud-radiation parameterizations. Four 78-day SCM numerical experiments were conducted for the atmospheric column overlying the COARE Intensive Flux Array. Each SCM experiment used a different cloud-radiation parameterization with a different representation of cloud microphysics. All the SCM experiments successfully reproduced most of the observed temporal variability in precipitation, cloud fraction, shortwave and longwave cloud forcing, and downwelling surface shortwave flux. The magnitude and temporal variability of the downward surface longwave flux was overestimated by all the SCM experiments. This bins is probably due to clouds forming too low in the model atmosphere. Time-averaged model results were used to examine the sensitivity of model performance to the differences between the four cloud-radiation parameterization packages. The SCM versions that calculated cloud amount as a function of cloud liquid water, instead of using a relative humidity-based cloud scheme, produced smaller amounts of both low and deep convective clouds. Additionally, larger high (cirrus) cloud emissivities were obtained with interactive cloud liquid water schemes than with the relative humidity-based scheme. Surprisingly. calculating cloud optical properties as a function of cloud liquid water amount, instead of parameterizing them based on temperature, humidity, and pressure, resulted in relatively little change in radiative fluxes. However. model radiative fluxes were sensitive to the specification of the effective cloud droplet radius. Optically thicker low clouds and optically thinner high clouds were produced when an interactive effective cloud droplet radius scheme was used instead of specifying a constant value. Comparison of model results to both surface and satellite observations revealed that model experiments that calculated cloud properties as a function of cloud liquid water produced more realistic cloud amounts and radiative fluxes. The most realistic vertical distribution of clouds was obtained from the SCM experiment that included the most complete representation of cloud microphysics. Due to the limitations of SCMs. the above conclusions are model dependent and need to be tested in a general circulation model.

Xie, SC, Xu KM, Cederwall RT, Bechtold P, Delgenio AD, Klein SA, Cripe DG, Ghan SJ, Gregory D, Iacobellis SF, Krueger SK, Lohmann U, Petch JC, Randall DA, Rotstayn LD, Somerville RCJ, Sud YC, Von Salzen K, Walker GK, Wolf A, Yio JJ, Zhang GJ, Zhang MG.  2002.  Intercomparison and evaluation of cumulus parametrizations under summertime midlatitude continental conditions. Quarterly Journal of the Royal Meteorological Society. 128:1095-1135.   10.1256/003590002320373229   AbstractWebsite

This study reports the Single-Column Model (SCM) part of the Atmospheric Radiation Measurement (ARM)/the Global Energy and Water Cycle Experiment (GEWEX) Cloud System Study (GCSS) joint SCM and Cloud-Resolving Model (CRM) Case 3 intercomparison study, with a focus on evaluation Of Cumulus parametrizations used in SCMs. Fifteen SCMs are evaluated under summertime midlatitude continental conditions using data collected at the ARM Southern Great Plains site during the summer 1997 Intensive Observing Period. Results from ten CRMs are also used to diagnose problems in the SCMs. It is shown that most SCMs can generally capture well the convective events that were well-developed within the SCM domain, while most of them have difficulties in simulating the occurrence of those convective events that only occurred within a small part of the domain. All models significantly underestimate the surface stratiform precipitation. A third of them produce large errors in surface precipitation and thermodynamic structures. Deficiencies in convective triggering mechanisms are thought to be one of the major reasons. Using a triggering mechanism that is based on the vertical integral of parcel buoyant energy without additional appropriate constraints results in overactive convection, which in turn leads to large systematic warm/dry biases in the troposphere. It is also shown that a non-penetrative convection scheme can underestimate the depth of instability for midlatitude convection, which leads to large systematic cold/moist biases in the troposphere. SCMs agree well quantitatively with CRMs in the updraught mass fluxes, while most models significantly underestimate the downdraught mass fluxes. Neglect of mesoscale updraught and downdraught mass fluxes in the SCMs contributes considerably to the discrepancies between the SCMs and the CRMs. In addition, uncertainties in the diagnosed mass fluxes in the CRMs and deficiencies with cumulus parametrizations are not negligible. Similar results are obtained in the sensitivity tests when different forcing approaches are used. Finally. sensitivity tests from an SCM indicate that its simulations can be greatly improved when its triggering mechanism and closure assumption are improved.

Razafimpanilo, H, Frouin R, Iacobellis SF, Somerville RCJ.  1995.  Methodology for estimating burned area from AVHRR reflectance data. Remote Sensing of Environment. 54:273-289.   10.1016/0034-4257(95)00154-9   AbstractWebsite

Two methods are described to determine burned area from Advanced Very High Resolution Radiometer (AVHRR) data. The first method, or the ''linear method,'' employs Channel 2 reflectance, R(2), and is based on the nearly linear relationship between the fraction of pixel burned, P, and R(2). The second method, or the ''nonlinear method,'' employs the Normalized Difference Vegetation Index (NDVI) derived from Channels 1 and 2 reflectances, and is based on the nonlinear relationship P=f(NDVI), a polynomial of order 2 in NDVI. The coefficients of the polynomial are parameterized as a function of the NDVI of the background before the fire event. Radiative transfer simulations indicate that the linear method, unlike the nonlinear method, must be applied to top-of-atmosphere reflectances that have been corrected for atmospheric influence. Sensitivity studies suggest that the methods are subject to some limitations. To avoid discontinuity problems, the original background (just before the fire) must be characterized by a Channel 2 reflectance above 0.07 and by a positive NDVI. To separate the useful signal from atmospheric effects, the fire scar must occupy at least 20% and 12% of the pixel area in the case of savanna and green vegetation (e.g., forest), respectively When applied to uniform pixels, the mean relative error on the fraction of area burned is about 20% for the linear method and 10% for the nonlinear method. The linear method gives better results for nonuniform pixels, but neither method can be used when the pixel contains low reflectance backgrounds (e.g., water).

Xu, KM, Zhang MH, Eitzen MA, Ghan SJ, Klein SA, Wu XQ, Xie SC, Branson M, Delgenio AD, Iacobellis SF, Khairoutdinov M, Lin WY, Lohmann U, Randall DA, Somerville RCJ, Sud YC, Walker GK, Wolf A, Yio JJ, Zhang JH.  2005.  Modeling springtime shallow frontal clouds with cloud-resolving and single-column models. Journal of Geophysical Research-Atmospheres. 110   10.1029/2004jd005153   AbstractWebsite

This modeling study compares the performance of eight single-column models (SCMs) and four cloud-resolving models (CRMs) in simulating shallow frontal cloud systems observed during a short period of the March 2000 Atmospheric Radiation Measurement (ARM) intensive operational period. Except for the passage of a cold front at the beginning of this period, frontal cloud systems are under the influence of an upper tropospheric ridge and are driven by a persistent frontogenesis over the Southern Great Plains and moisture transport from the northwestern part of the Gulf of Mexico. This study emphasizes quantitative comparisons among the model simulations and with the ARM data, focusing on a 27-hour period when only shallow frontal clouds were observed. All CRMs and SCMs simulate clouds in the observed shallow cloud layer. Most SCMs also produce clouds in the middle and upper troposphere, while none of the CRMs produce any clouds there. One possible cause for this is the decoupling between cloud condensate and cloud fraction in nearly all SCM parameterizations. Another possible cause is the weak upper tropospheric subsidence that has been averaged over both descending and ascending regions. Significantly different cloud amounts and cloud microphysical properties are found in the model simulations. All CRMs and most SCMs underestimate shallow clouds in the lowest 125 hPa near the surface, but most SCMs overestimate the cloud amount above this layer. These results are related to the detailed formulations of cloud microphysical processes and fractional cloud parameterizations in the SCMs, and possibly to the dynamical framework and two-dimensional configuration of the CRMs. Although two of the CRMs with anelastic dynamical frameworks simulate the shallow frontal clouds much better than the SCMs, the CRMs do not necessarily perform much better than the SCMs for the entire period when deep and shallow frontal clouds are present.

Iacobellis, SF, Frouin R, Razafimpanilo H, Somerville RCJ, Piper SC.  1994.  North African savanna fires and atmospheric carbon dioxide. Journal of Geophysical Research-Atmospheres. 99:8321-8334.   10.1029/93jd03339   AbstractWebsite

The effect of north African savanna fires on atmospheric CO2 is investigated using a tracer transport model. The model uses winds from operational numerical weather prediction analyses and provides CO2 Concentrations as a function of space and time. After a spin-up period of several years, biomass-burning sources are added, and model experiments are run for an additional year, utilizing various estimates of CO2 sources. The various model experiments show that biomass burning in the north African savannas significantly affects CO2 concentrations in South America. The effect is more pronounced during the period from January through March, when biomass burning in South America is almost nonexistent. During this period, atmospheric CO2 concentrations in parts of South America typically may increase by 0.5 to 0.75 ppm at 970 mbar, the average pressure of the lowest model layer. These figures are above the probable uncertainty level, as model runs with biomass-burning sources estimated from independent studies using distinct data sets and techniques indicate. From May through September, when severe biomass burning occurs in South America, the effect of north African savanna fires over South America has become generally small at 970 mbar, but north of the equator it may be of the same magnitude or larger than the effect of South American fires. The CO2 concentration increase in the extreme northern and southern portions of South America, however, is mostly due to southern African fires, whose effect may be 2-3 times larger than the effect of South American fires at 970 mbar. Even in the central part of the continent, where local biomass-burning emissions are maximum, southern African fires contribute to at least 15% of the CO2 concentration increase at 970 mbar. At higher levels in the atmosphere, less CO2 emitted by north African savanna fires reaches South America, and at 100 mbar no significant amount of CO2 is transported across the Atlantic Ocean. The vertical structure of the CO2 concentration increase due to biomass burning differs substantially, depending on whether sources are local or remote. A prominent maximum Of CO2 concentration increase in the lower layers characterizes the effect of local sources, whereas a more homogenous profile of CO2 concentration increase characterizes the effect of remote sources. The results demonstrate the strong remote effects of African biomass burning which, owing to the general circulation of the atmosphere, are felt as far away as South America.

McFarquhar, GM, Iacobellis S, Somerville RCJ.  2003.  SCM simulations of tropical ice clouds using observationally based parameterizations of microphysics. Journal of Climate. 16:1643-1664.   10.1175/1520-0442(2003)016<1643:ssotic>;2   AbstractWebsite

A new bulk parameterization of the dependence of ice cloud effective radius (r(e)) on ice water content (IWC) is developed using in situ observations of the size and shape of ice crystals in tropical anvils. This work extends previous parameterizations because information about the number, size, and shape of ice crystals with diameters smaller than 100 m m is included and in that a range of possible fit coefficients, rather than single values, is given to reflect the fact that r(e) can vary significantly about its mean parameterized value. The parameterization is implemented in the Scripps single column model (SCM), and simulations of tropical clouds over the Atmospheric Radiation Measurement ( ARM) program's tropical western Pacific (TWP) site and over the Tropical Ocean Global Atmosphere Coupled Ocean-Atmosphere Response Experiment (TOGA COARE) domain are conducted. Sensitivity studies determine how the range of possible fit coefficients, which reflects the uncertainty in the parameterization of r(e), relates to uncertainties in modeled cloud radiative forcings (CRFs). When r(e) is chosen one or two standard deviations higher or lower than the mean parameterized value, temporally averaged shortwave CRFs can differ by up to 17.7 W m(-2) from that value estimated from the mean parameterized r(e), the difference depending on the time period and location; differences in longwave CRFs are substantially less. When other uncertainties in the parameterization are accounted for, such as those based on the observed numbers of smaller crystals, CRFs can differ by up to 25 W m(-2) from that determined by the base parameterization. When r(e) is randomly chosen for each simulation time within one or two standard deviations of the most likely r(e) for that IWC, shortwave CRFs can still differ from that of the base simulation by up to 13.9 W m(-2), with an enhancement of shortwave reflection of up to 4.9 W m(-2) observed on average. Therefore, the average of a series of such simulations may not equal a simulation of average conditions, a finding that may have important ramifications. Both interactive simulations, where changes in cloud heating rates feed back upon predicted cloud masses, and noninteractive simulations, where changes in heating rate do not feed back upon cloud mass, are performed in order to determine how and why different parameterizations affect the CRFs. It is shown that differences in longwave heating rates, associated with different versions of the parameterization, alter the mass of ice and liquid water produced at various levels, this change in cloud mass in turn affects the CRF. This change can either amplify or reduce the change in CRF associated with the more direct effect of varying the r(e) parameterization, namely, that smaller particles reflect more shortwave radiation given the same mass content. The amount of liquid water present in low clouds is an important indicator of whether changing ice cloud microphysical properties will have an important effect on CRF.

Lane, DE, Somerville RCJ, Iacobellis SF.  2000.  Sensitivity of cloud and radiation parameterizations to changes in vertical resolution. Journal of Climate. 13:915-922.   10.1175/1520-0442(2000)013<0915:socarp>;2   AbstractWebsite

The importance of vertical resolution to the parameterization of cloud-radiation processes in climate models is examined. Using a one-dimensional single-column model containing a typical suite of physical parameterizations, the authors test 12 different vertical resolutions, ranging from 16 to 60 layers. The model products are evaluated against observational data taken during three intensive observation periods from the Atmospheric Radiation Measurement Program. The simulated values of cloud-radiation variables display a marked sensitivity to changes in vertical resolution. This sensitivity is apparent in all the model variables examined. The cloud fraction varies typically by approximately 10% over the range of resolutions tested, a substantial amount when compared to the typical observed values of about 50%. The outgoing longwave radiation typically changes by approximately 10-20 W m(-2) as resolution is varied, which is of the order of 5%-10% of the observed value. The downwelling shortwave radiation change is somewhat smaller but is still significant. Furthermore, the model results have not converged even at a resolution of 60 layers, and there are systematic differences between model results and observations.

Iacobellis, SF, McFarquhar GM, Mitchell DL, Somerville RCJ.  2003.  The sensitivity of radiative fluxes to parameterized cloud microphysics. Journal of Climate. 16:2979-2996.   10.1175/1520-0442(2003)016<2979:tsorft>;2   AbstractWebsite

The sensitivity of modeled radiative fluxes to the specification of cloud microphysical parameterizations of effective radius and fallout are investigated using a single-column model and measurements from the Atmospheric Radiation Measurement (ARM) Program. The single-column model was run with data for the 3-month period of June - August 2000 at the ARM Southern Great Plains site forced with operational numerical weather prediction data. Several different packages of cloud microphysical parameterizations were used in the single-column model. The temporal evolution of modeled cloud amount as well as surface radiative fluxes from a control run compare well with ARM measurements. Mean ice particle fall speeds varied significantly with respect to the assumed ice particle habit. As particle fall speeds increased, the overall cloud fraction, cloud height, and grid-averaged ice water path decreased. The outgoing longwave radiation (OLR) differs by up to 4 W m(-2) over the range of fall speeds examined, while shortwave fluxes varied little as most of the changes in cloud properties occurred at times of minimal solar radiation. Model results indicate that surface and top-of-atmosphere radiative fluxes are sensitive to the scheme used to specify the ice particle effective radius. On the seasonal timescale this sensitivity is on the order of 4 W m(-2) and on the daily timescale can be as large as 32 W m(-2). A conclusive statement as to which microphysical scheme is performing best is not achievable until cloud microphysical measurements include an accurate representation of small ice particles. The modeled variance of the ice particle effective radius at any given height in the model is considerably smaller than that suggested by measurements. Model results indicate that this underestimation of the ice particle effective radius variance can alter the seasonal mean top-of-atmosphere radiative fluxes by up to 5 W m(-2) and the mean longwave cooling rate by up to 0.2degrees K day(-1) near the location of maximum cloud amount. These seemingly modest flux sensitivities may have important implications for numerical climate simulations. These numerical experiments and observational comparisons have provided valuable physical insight into ice cloud - radiation physics and also into the mechanisms through which contemporary cloud microphysical parameterizations interact with climate model radiation schemes. In particular, the results demonstrate the importance of the smaller ice particles and emphasize the critical role played by not only the average particle size and shape but also the width of the ice particle effective radius distribution about its mean. In fact, the results show that this variability in particle size can sometimes play a greater role in cloud - radiation interactions than the more obvious variations in cloud amount due to changes in ice particle fall speed.

Xie, SC, Zhang MH, Branson M, Cederwall RT, Delgenio AD, Eitzen ZA, Ghan SJ, Iacobellis SF, Johnson KL, Khairoutdinov M, Klein SA, Krueger SK, Lin WY, Lohmann U, Miller MA, Randall DA, Somerville RCJ, Sud YC, Walker GK, Wolf A, Wu XQ, Xu KM, Yio JJ, Zhang G, Zhang JH.  2005.  Simulations of midlatitude frontal clouds by single-column and cloud-resolving models during the Atmospheric Radiation Measurement March 2000 cloud intensive operational period. Journal of Geophysical Research-Atmospheres. 110   10.1029/2004jd005119   AbstractWebsite

[1] This study quantitatively evaluates the overall performance of nine single-column models (SCMs) and four cloud-resolving models (CRMs) in simulating a strong midlatitude frontal cloud system taken from the spring 2000 Cloud Intensive Observational Period at the Atmospheric Radiation Measurement ( ARM) Southern Great Plains site. The evaluation data are an analysis product of constrained variational analysis of the ARM observations and the cloud data collected from the ARM ground active remote sensors (i.e., cloud radar, lidar, and laser ceilometers) and satellite retrievals. Both the selected SCMs and CRMs can typically capture the bulk characteristics of the frontal system and the frontal precipitation. However, there are significant differences in detailed structures of the frontal clouds. Both CRMs and SCMs overestimate high thin cirrus clouds before the main frontal passage. During the passage of a front with strong upward motion, CRMs underestimate middle and low clouds while SCMs overestimate clouds at the levels above 765 hPa. All CRMs and some SCMs also underestimated the middle clouds after the frontal passage. There are also large differences in the model simulations of cloud condensates owing to differences in parameterizations; however, the differences among intercompared models are smaller in the CRMs than the SCMs. In general, the CRM-simulated cloud water and ice are comparable with observations, while most SCMs underestimated cloud water. SCMs show huge biases varying from large overestimates to equally large underestimates of cloud ice. Many of these model biases could be traced to the lack of subgrid-scale dynamical structure in the applied forcing fields and the lack of organized mesoscale hydrometeor advections. Other potential reasons for these model errors are also discussed in the paper.

Randall, DA, Xu KM, Somerville RJC, Iacobellis S.  1996.  Single-column models and cloud ensemble models as links between observations and climate models. Journal of Climate. 9:1683-1697.   10.1175/1520-0442(1996)009<1683:scmace>;2   AbstractWebsite

Among the methods that have been devised to test physical parameterizations used in general circulation models, one of the most promising involves the use of field data together with single-column models (SCMs) and/or cloud ensemble models. Here the authors briefly discuss the data requirements of such models and then give several examples of their use. Emphasis is on parameterizations of convection and cloud amount.

Somerville, RCJ, Iacobellis SF.  1999.  Single-column models, ARM observations, and GCM cloud-radiation schemes. Physics and Chemistry of the Earth Part B-Hydrology Oceans and Atmosphere. 24:733-740.   10.1016/s1464-1909(99)00074-x   AbstractWebsite

Among the most serious sources of uncertainty in current general circulation models (GCMs) is the treatment of clouds and cloud-radiation interactions. We have used a single-column model (SCM) diagnostically to evaluate parameterizations against observations from the Atmospheric Radiation Measurement (ARM) Program. We find that schemes with explicit cloud water budgets and interactive radiative properties are potentially capable of matching observational data closely. In our SCM, using an interactive cloud droplet radius decreases the cloud optical thickness and cloud infrared emittance of high clouds, which acts to increase the downwelling surface shortwave flux and the outgoing longwave radiation. However, it is difficult to evaluate the realism of the vertical distribution of model-produced cloud extinction, cloud emittance, cloud liquid water content and effective cloud droplet radius until high-quality observations of these quantities become more widely available. We also find that in the SCM, cloud parameterizations often underestimate the observed cloud amount, and that ARM observations indicate the presence of clouds while the corresponding maximum relative humidity is less than 80%. This implies that the underlying concept of a critical gridpoint relative humidity of about 80% for cloud formation, as used in many GCM cloud parameterizations, may need to be reexamined. (C) 1999 Elsevier Science Ltd. All rights reserved.