Johannes Schmetz

Towards a surface radiation climatology: Retrieval of downward irradiances from satellites

Abstract Methods are reviewed for retrieving the downward shortwave (0.3–4 μm) and longwave (4–100 μm) irradiances at the earths surface from satellites. Emphasis is placed on elucidating the physical aspects relevant to the satellite retrieval. For the shortwave irradiance an example of a retrieval is presented. The shortwave retrieval is facilitated by a close linear coupling between the reflected radiance field at the top of the atmosphere and the surface irradiance. A linear relationship between planetary albedo and surface irradiance does also account properly for cloud absorption, since cloud absorption and albedo are linearly related. In the longwave the retrieval is more difficult since only atmospheric window radiances at the top of the atmosphere can bear information on the near-surface radiation field. For the remainder of the longwave spectrum the radiation regimes at the top of the atmosphere and at the surface are decoupled. More than 80% of the clear-sky longwave flux reaching the surface is emitted within the lowest 500 m of the atmosphere. In cloudy conditions the radiation fields at the surface and at the top of the atmosphere are entirely decoupled. Cloud contributions to the surface irradiance are important within the atmospheric window (8–13 μm) and the relative contribution increases in drier climates. Summaries are presented of various techniques devised for both the solar and longwave surface irradiances. A compilation of reported standard errors of shortwave techniques in comparison with ground measurements yields median values of about 5% and 10% for monthly and daily mean values, respectively. Standard errors for the longwave are of the order of 10–25 W m −2 . Reported biases are typically of the order of 5 W m −2 . For the shortwave retrieval there are fairly good prospects to obtain monthly mean estimates with the requested accuracy of about 10 W m −2 over regional scale areas. The inherent problems of the longwave still entails improvements. The requested retrieval accuracy may be reached with advanced techniques for estimating cloud base height and with the exploitation of correlative data, such as the analysis for numerical weather prediction of fields of temperature and humidity. The use of such data should also be advantageous to physical methods for the shortwave retrieval. Validation studies are compromised by the different nature of single spot surface measurements and area covering satellite retrievals. For physical retrievals employing radiative transfer codes it is recommended to test models against a defined standard.

Upper tropospheric humidity observations from Meteosat compared with short‐term forecast fields

Monthly mean brightness temperature observations from the water vapor channel (WV: 5.7–7.1 μ m) aboard the geostationary weather satellite METEOSAT-4 are presented for July 1992. The WV channel is sensitive to the atmospheric column water vapor in the upper troposphere above 600 to 500 hPa. The WV channel has been recalibrated with a radiative transfer model using quality controlled radiosondes. The observations are compared with brightness temperature calculations based on short-term forecast fields (12 and 24 h) of the general circulation model of the European Centre for Medium Range Weather Forecasts (ECMWF). While the model humidity field closely resembles the observed large-scale distribution it is too moist in the subtropics and slightly too dry in areas associated with the Inter Tropical Convergence Zone (ITCZ). A simple method is suggested that quantifies the moisture bias in the general circulation model by adjusting the model based brightness temperatures.

Retrieval of surface radiation fluxes from satellite data

Abstract The retrieval of the downward solar (0.3–4 μm) and longwave (4–100 μm) irradiances and the direct retrieval of the solar net flux at the surface from satellite observations are discussed. A close linear coupling between the solar radiation field observed by the satellite and the surface radiation field enables surface insolation estimates with standard errors of about 5% for monthly mean values in comparison with surface measurements. It is pointed out that the net solar flux density at the surface can be calculated directly from the net at the top of the atmosphere with an adequate estimate of absorption in the atmosphere. In the longwave the satellite estimates are more difficult since the downwelling longwave irradiance at the Earths surface is largely decoupled from the radiation field at the top of the atmosphere. The clear-sky longwave irradiance is basically determined by the near-surface temperature and humidity field. Clouds with low bases are very effective for increasing the downward longwave flux. The accuracy of the present longwave retrievals is not yet adequate for the derivation of a global climatology while shortwave retrievals provide useful results.

Operational cloud motion winds from meteosat and the use of cirrus clouds as tracers

Abstract The operational method for deriving cloud motion winds from METEOSAT infrared (IR: 10.5–12.5μm) images and recent changes are described. The wind extraction works fully automatically and is followed by a manual quality control. At present about 3000 wind vectors per day are produced with four production runs. Improvements to the height assignment of a wind vector and radiance filtering techniques preceding the cloud tracking have ameliorated the METEOSAT winds significantly. The major shortcoming of cloud motion winds is their tendency to underestimate the high speed wind field around jet stream areas, which typically are characterized by extended cirrus cloud. The second part of the paper deals with the use of cirrus clouds as tracers for the wind field. The problem of the height assignment of wind vectors and the question of whether satellite deduced displacements from cirrus clouds do represent a single level wind are discussed.

Relative effects of solar and infrared radiative forcing in a mesoscale model

A new efficient parameterization scheme for solar short-wave radiative heating, as a component of the net radiative effects in the atmosphere, is tested in a three-dimensional mesoscale model. This model is designed with moist convective processes in mind, so that the radiative parameterization (solar plus thermal infrared) are interactive with the cloud field. Previous work by the authors with only an infrared scheme has demonstrated that cloud-radiation interactions are characterized by strong cloud-top cooling, leading to upper cloud-layer destabilization. The effects of including solar heating are to modulate the strength of the strong infrared cooling, thereby leading to weaker interactions between clouds, radiation, and mesoscale fields. The present study shows that even on the mesoscale and for relatively short time-spans, radiative processes in the presence of clouds are not negligible.As a further step, a simple fractional cloud cover parameterization is introduced and the model response is compared with results omitting this parameterization.

A three-dimensional study of mesoscale model response to radiative forcing

An infrared radiation parameterization has been applied to a detailed three-dimensional mesoscale model in order to determine whether radiative forcing significantly affects mesoscale atmospheric processes. By taking into account water vapor, liquid water, and carbon dioxide absorption, the scheme differentiates between cloud and clear air regions. The parametric model is presented, along with an overview of the associated mesoscale model.Comparisons between a control run in which only a uniform cooling rate of l K day−1 is specified, and runs with the infrared scheme are made for 12-hr simulations. The major feature of the radiative forcing is seen to be strong cloud-top cooling. This leads to enhanced destabilization of the upper cloud layer, which in turn results in faster growth of clouds (and which extend to higher levels) than in the control experiment. The deeper clouds force a more vigorous secondary circulation, in which thermodynamic feedbacks between clouds and their environment are substantially stronger than in the case with only a constant cooling rate. This confirms findings made in previous studies undertaken in small-scale numerical models. The discussion also focuses upon a simulation in which the cloud-top infrared cooling has been smoothed out over neighboring vertical levels, in order to represent a cloud-top height distribution crudely. The results indicate that although the absolute values of cloud-top cooling are reduced with respect to the unfiltered case, the fact that cooling extends even higher than previously predicted leads to the formation of thicker, more vigorous clouds. These clouds interact more intensely with their environment than in the unfiltered situation, thereby considerably modifying the mesoscale atmosphere.

On the retrieval of surface radiation budget components from satellites

Abstract The retrieval of the surface radiation budget components in both the solar (0.3–4 μm) and longwave (4–100 μm) parts of the spectrum is discussed. A close linear coupling between the solar radiation field observed by the satellite and the surface radiation field facilitates the retrieval for solar radiation. Simple equations for a retrieval of the downwelling solar irradiance, the surface albedo and the net solar flux can be derived from the local solar radiative energy budget of the atmosphere-surface system. Generally the solar retrievals are advanced and provide useful results for climatological studies. In the longwave the satellite estimates are more difficult since the downwelling longwave irradiance at the Earths surface is largely decoupled from the radiation field at the top of the atmosphere. The clear-sky longwave irradiance is basically determined by the near-surface temperature and humidity field. Clouds with low bases are very effective for increasing the downward longwave flux.

Cloud motion winds from METEOSAT: performance of an operational system

Abstract Wind fields have been estimated from the displacement of clouds in successive METEOSAT infrared (IR: 10.5–12.5 μm) images for more than a decade. The derivation of cloud motion winds (hereafter: CMW) workss fully automatically, man-machine interaction is only performed as the very last step of manual quality control. At present about 3000 wind vectors per day are produced with four production runs and disseminated via the Global Telecommunication System (GTS) mainly for use in the analysis for numerical weather prediction. In this paper the improvements to the CMW derivation from METEOSAT images are described. In particular the height assignment of a wind vector and radiance filtering techniques preceding the cloud tracking have ameliorated the errors in METEOSAT winds significantly. The low speed bias of high level CMWs (

Cloud observations from METEOSAT and the inference of winds

Abstract At the European Space Operations Centre (ESOC) the raw image data from the geostationary meteorological satellite METEOSAT are received, preprocessed and operational products are derived. Most products directly or indirectly make use of cloud information retrieved by a multispectral image analysis. The purpose of the paper is to describe the cloud retrieval and the subsequent use of the cloud information in the derivation of other products. In particular we discuss the production of cloud motion winds (CMW) based on an automatic cloud tracking and the estimation of the outgoing longwave radiation (OLR) from a condensed data set which is the direct output of the multispectral analysis. Results are presented for the monthly mean OLR and the longwave cloud radiative forcing for July 1983.

Wind fields and upper tropospheric humidity from satellite images in the 6.3 μm band

This paper presents results of a pilot study investigating the relationship between the monthly mean fields of wind and humidity in the upper troposphere. The wind fields are derived from successive METEOSAT images in the water vapour channel (WV: 5.7 – 7.1 μm). The upper tropospheric relative humidity (UTH) is inferred from water vapour brightness temperatures with a modified regression scheme that explicitly considers temperature lapse rate. n nQuantitative information on the large scale circulation in the upper troposphere can be derived from WV wind fields. The monthly mean wind field of January 1992 is used to estimate the large scale divergence, which ranges between about −8.10−6 s−1 and 8.10−6 s−1 on a scale of about 1500 km. The spatial pattern of the UTH field closely resembles the divergence of the wind field suggesting that UTH is principally determined by large scale circulation. Implications for weather prediction and climate modeling are discussed.