Satellite Detection Of Multi-layered Clouds And Associated Radiative Effects

Multi-layered clouds can significantly affect the Earth-atmosphere system's energy budget,the atmospheric circulation and the water vapor cycle by changing the atmospheric temperature,atmospheric radiative heating rate/cooling rate and latent heat.Given the limitation of the detective methods and the scarcity of the ground observation data,the study of multi-layered clouds' occurrence frequencies,associated radiation effects as well as the interactions between cloud layers,however,are insufficient.Moreover,the unreasonable representation of the cloud overlap which is closely related to multi-layered clouds in climate model is also an important source of uncertainties in prediction of climate change.Based on their advantages in detecting cloud vertical structures,active satellites have developed several sets of remote sensing products in recent years,which are widely used in different research areas.Therefore,in this paper we systematically carry out research work on multi-layered clouds by mainly using these remote sensing products from active satellites.Research results are able not only to provide the necessary research basis for further understanding the global cloud radiation feedback mechanism,but also to improve the cloud overlap parameterization schemes in models based on the observation.Specific contents and major conclusions are summarized as follows.(1)We assessed the downwelling radiative fluxes at the surface(DRSs)in the radiative fluxes products of active satellites(2B-FLXHR-LIDAR)by using coincident surface measurements obtained from six Atmospheric Radiation Measurement(ARM)program sites and one China site called the Semi-arid Climate and Environment Observatory of Lanzhou University(SACOL).The statistical results show that,on the one hand,the shortwave DRS retrieved by active satellites agrees well with surface results examined by sites in the situation of clear sky with biases about 20~50 W/m~2.On the other hand,the shortwave DRS retrieved by active satellites is significantly higher than surface results examined by sites under cloudy conditions with biases about 70~180 W/m~2.The phenomena demonstrated that the shortwave DRS retrieved by active satellites cannot fully count the cloud impacts and has relatively large uncertainties.In contrast,the longwave DRSs during both day and night retrieved by active satellites are in good agreementwith the surface observations for both clear and cloudy sky conditions.Another finding is that the impacts of both temporal and spatial scales on the DRS biases are small but cloud phases demonstrate significant impacts to the DRS biases.(2)Based on the radiative fluxes product(2B-FLXHR-LIDAR)and the cloud classification product(2B-CLDCLASS-LIDAR)from active satellites from 2007 to 2010,we presented a quantitative analysis of cloud fraction,weighted cloud radiative forcing(CRF),and cloud radiative heating rate(CRH)of single-layered clouds(SLC)and multi-layered clouds(MLC),and their differences on the global scale.The statistical results show that the globally averaged cloud fraction of the MLC(24.9%),which is primarily prevalent in equatorial regions,is smaller than that of the SLC(46.6%).The MLC contributes approximately 40.2%,42.4%,and 57% to the net weighted CRF at the top and bottom of the atmosphere,and in the atmosphere,respectively.Moreover,the MLC exhibits distinct differences to the SLC in terms of the CRH.The shortwave CRH of the SLC(MLC)reaches a heating peak at 9.75(7.5)km,with a value of 0.35(0.60)K/d,and the differences between SLC and MLC transform from positive to negative with increasing altitude.Meanwhile,the longwave CRH of the SLC(MLC)reaches a cooling peak at 2(8)km,with a value of-0.45(-0.42)K/d,and the differences transform from negative to positive with increasing altitude.To sum up,the net CRH differences between SLC and MLC are negative below 7.5 km.(3)By matching the cloud classification product(2B-CLDCLASS-LIDAR)from active satellites with the CERES instantaneous radiation data(CERES-SSF)from passive satellites from~2007 to 2010,we evaluated cloud characteristics of the single-layered cloud types and associated CRF at at the top of atmosphere(TOA),and further statistically studied impacts of interactions between overlapping high clouds and underlying cloud types from two-layered cloud types on macro-and micro-physical properties of each cloud layer and the radiative balance at TOA.Results indicate that absolute values of the shortwave(or longwave)CRF for all single-layered cloud types increase with the growing of the cloud optical thickness,and absolute values of their correlation coefficients are not less than 0.78(or 0.84).When high clouds overlap other different cloud types,the radiative interactions between the two cloud layers are able to significantly change cloud characteristics of both upper and lower cloud layers,demonstrating that overlying high clouds reveal more opaque and thicker cloud bodies than single-layered high clouds,while underlying cloud types exhibit more transparent and thinner cloud bodies than their single-layered status.Comparing single-layered cloud types and their two-layered statuses overlapping by high clouds in terms of the net CRF,we found that all single-layered cloud types reveal stronger net cooling effects on the Earth-atmosphere system,except for the single-layered cumulus.(4)Through taking advantage of the 2B-GEOPROF-LIDAR product from active satellites,the ECMWF-AUX auxiliary product and ERA-Interim reanalysis from 2007 to 2010,we retrieved overlap parameters of clouds over the Tibetan Plateau(TP)and further assessed impacts of dynamic parameters,such as wind shear and the vertical gradient of saturation equivalent potential temperature on the cloud overlap parameter.The statistical results reveal that the characteristics of cloud overlap for non-continuous cloud can be depicted well by the random overlap scheme.The cloud overlap parameter for continuous clouds,however,tends to converge from maximum overlap to random overlap as the cloud layer separation distance increasing.The cloud overlap parameter become especially negative as the cloud layer separation distance exceeds 1.5 km,indicating that clouds over TP begin to tend to minimum overlap scheme.Our statistical results show that the sample size of negative cloud overlap parameter in the TP region accounts for 41%of the total sample size,which is significantly different from globally averaged results proposed by earlier studies.The phenomenon is closely related to the unique atmospheric dynamic characteristics of the TP.We found that atmospheric stability and wind shear significantly affect the overlap characteristic of clouds over the TP,and further parameterize the decorrelation index which is related to cloud overlap into a function of atmospheric stability and wind shear.The comparisons between new parametric formulas and other formulas in terms of the simulated total cloud cover demonstrate that total cloud covers predicted by new parametric formulas have small deviations.Therefore,our study suggests that the cloud overlap scheme in models should consider effects of atmospheric dynamic parameters for improving the simulation of regional total cloud cover.