Abstract

A semistatistical retrieval technique is presented to derive humidity profiles over the oceans from passive microwave measurements. The procedure is based upon the vertical empirical orthogonal functions (EOFs) of the specific humidity extracted from a large sample of radiosonde measurements over the North Atlantic Ocean during the period from April to October 1979 (FGGE-year). The North Atlantic is divided into seven regions and the EOF-analysis is carried out for each region separately. The first three eigenvectors of the EOF-expansion explain up to 90% of the total variability within each region and it is shown that they are statistically significant and stable. The eigenvectors of the first order mainly describe variations of the total precipitable water (W), while the second and third-order EOFs are related to the ratio W_G /W, with W_G as the precipitable water of the planetary boundary layer, and the sea surface temperature (SST), respectively. This fact is used to develop a technique for the estimation of atmospheric moisture profiles assuming that W, W_G , and SST can be retrieved from satellite observations.

A comparison of the SMMR (Nimbus-7) derived humidity profiles using only W as input data yields a retrieval accuracy of 0.9 g kg⁻¹ in the surface layer and 1.4 g kg⁻¹ at 800 hPa, which corresponds roughly with the top of the PBL. Above this layer the retrieval error falls off rapidly. Case studies demonstrate the capability of the algorithm to resolve typical structures of the humidity field, e.g., synoptic scale disturbances in midlatitudes or the intertropical convergence zone (ITCZ) in the tropics.

Abstract

Multiple-scattering effects as sensed by radars in configurations useful in the context of the Global Precipitation Mission (GPM) are evaluated for a range of meteorological profiles extracted from four different cloud-resolving model simulations. The multiple-scattering effects are characterized in terms of both the reflectivity enhancement and the linear depolarization ratio. When considering the copolarized reflectivity in spaceborne configurations, the multiple-scattering enhancement becomes a real issue for K_a-band radars, though it is generally negligible at the K_u band, except in meteorologically important situations such as when high rain rates and a considerable amount of ice are present aloft. At K_a band it can reach tens of decibels when systems of heavy cold rain are considered, that is, profiles that include rain layers with high-density ice particles aloft. On the other hand, particularly at 35 GHz, high values of the linear depolarization ratio are predicted even in airborne configurations because of multiple-scattering effects. This result should allow the observation of these features in field campaigns.

Abstract

A numerical model based on the Monte Carlo solution of the vector radiative transfer equation has been adopted to simulate radar signals. The model accounts for general radar configurations such as airborne/spaceborne/ground based and monostatic/bistatic and includes the polarization and the antenna pattern as particularly relevant features. Except for contributions from the backscattering enhancement, the model is particularly suitable for evaluating multiple-scattering effects. It has been validated against some analytical methods that provide solutions for the first and second order of scattering of the copolar intensity for pencil-beam/Gaussian antennas in the transmitting/receiving segment. The model has been applied to evaluate the multiple scattering when penetrating inside a uniform hydrometeor layer. In particular, the impact of the phase function, the range-dependent scattering optical thickness, and the effects of the antenna footprint are considered.

Abstract

A method is introduced to derive cloud effects on the earth radiation budget. The ISCCP Cl cloud data for daylight cases are used in combination with a radiative transfer model to estimate the outgoing broadband radiative fluxes at the top of the atmosphere. Two tests are performed: the modeled narrowband filtered radiances are verified against the ISCCP satellite observation data (internal tests) and the modeled broadband fluxes are compared against ERBE data (external tests). After successful completion of the tests, the reflected solar (OSR) and the outgoing longwave (OLR) radiation for each of the 35 ISCCP cloud classes are determined. Results are shown for a reduced set of nine cloud classes (three bins for cloud top height, three bins for the optical thickness).

The global monthly and annual means of albedo and greenhouse effect are given for one month of each season (April, July, October 1985; January 1986). For the global annual mean cloud radiative cooling of the earth in the shortwave spectral range dominates over cloud radiative warming in the longwave spectral range by a factor of 2. Seasonal variations of the net effect are mainly due to changes in the shortwave cloud-forcing component [−43.9 to −52.9 W m⁻² (shortwave), 22.5 to 25.1 W m⁻² (longwave)]. The mean sensitivity of the net radiation budget to changes in cloud amount attains its minimum of −29.4 W m⁻² per 100% change in cloud cover in April and its maximum of −47.7 W m⁻² per 100% change in cloud cover in January. This is also mainly due to the solar component. A comparison with results of other studies are shown.

The effects of the nine cloud types are analyzed in detail. Cooling, warming, and radiatively neutral cloud types could be distinguished. Low stratus and midlevel nimbostratus clouds contribute about 80% to the total cooling of all clouds. Their effect on the OLR, however, is small (less than 20%) but they contribute almost 50% to the OSR. The deep convective clouds produce the largest single effect for both longwave (8.8 W m⁻²) and shortwave (−1 3 W m⁻²) forcing but their net effect is small. Only cirrus clouds have net warming effects.

Abstract

A method for combining ground-based passive microwave radiometer retrievals of integrated liquid water (LWP), radar reflectivity profiles (Z), and statistics of a cloud model is proposed for deriving cloud liquid water profiles (LWC). A dynamic cloud model is used to determine Z–LWC relations and their errors as functions of height above cloud base. The cloud model is also used to develop an LWP algorithm based on simulations of brightness temperatures of a 20–30-GHz radiometer. For the retrieval of LWC, the radar determined Z profile, the passive microwave retrieved LWP, and a model climatology are combined by an inverse error covariance weighting method. Model studies indicate that LWC retrievals with this method result in rms errors that are about 10%–20% smaller in comparison to a conventional LWC algorithm, which constrains the LWC profile exactly to the measured LWP. According to the new algorithm, errors in the range of 30%–60% are to be anticipated when profiling LWC. The algorithm is applied to a time series measurement of a stratocumulus layer at GKSS in Geesthacht, Germany. The GKSS 95-GHz cloud radar, a 20–30-GHz microwave radiometer, and a laser ceilometer were collocated within a 5-m radius and operated continuously during the measurement period. The laser ceilometer was used to confirm the presence of drizzle-sized drops.

Abstract

This paper evaluates several daily precipitation products over western and central Europe, identifies and documents their respective strengths and shortcomings, and relates these to uncertainties associated with each of the products. We analyze one gauge-based, three satellite-based, and two reanalysis-based products using high-density rain gauge observations as reference. First, we assess spatial patterns and frequency distributions using aggregated statistics. Then, we determine the skill of precipitation event detection from these products with a focus on extremes, using temporally and spatially matched pairs of precipitation estimates. The results show that the quality of the datasets largely depends on the region, season, and precipitation characteristic addressed. The satellite and the reanalysis precipitation products are found to have difficulties in accurately representing precipitation frequency with local overestimations of more than 40%, which occur mostly in dry regions (all products) as well as along coastlines and over cold/frozen surfaces (satellite-based products). The frequency distributions of wet-day intensities are generally well reproduced by all products. Concerning the frequency distributions of wet-spell durations, the satellite-based products are found to have clear deficiencies for maritime-influenced precipitation regimes. Moreover, the analysis of the detection of extreme precipitation events reveals that none of the non-station-based datasets shows skill at the shortest temporal and spatial scales (1 day, 0.25°), but at and above the 3-day and 1.25° scale the products start to exhibit skill over large parts of the domain. Added value compared to coarser-resolution global benchmark products is found both for reanalysis and satellite-based products.

Abstract

A two-way coupling of the operational mesoscale weather prediction model known as Lokal Modell (LM; German Weather Service) with the land surface hydrologic “TOPMODEL”-Based Land Surface–Atmosphere Transfer Scheme (TOPLATS; Princeton University) has been carried out to investigate the influence of a “state-of-the-art” land surface hydrologic model on the predicted local weather. Two case studies are presented that quantify the influence of the combined modeling system on the turbulent fluxes and boundary layer structure and on the formation of precipitation. The model results are compared with ground-based measurements of turbulent fluxes, boundary layer structure, and precipitation. Furthermore, whether the initialization of the original LM with more realistic soil moisture fields would be sufficient to improve the weather forecast is investigated. The results of the two case studies show that, when compared with measurements, the two-way coupled modeling system using TOPLATS improves the predicted energy fluxes and rain amount in comparison with predictions from the original LM. The initialization of the LM just using soil moisture fields based on TOPLATS does not result in an improvement of the local weather forecast: although the simulation of the sensible and latent heat fluxes is improved, the representation of the boundary layer structure is not captured well. In the original LM, the surface processes are not modeled in sufficient detail, which resulted in significant overprediction of precipitation for one case study. The main reason for the improved performance of the two-way coupled modeling system on the basis of TOPLATS probably is the more accurate representation of vegetation and soil hydrologic processes. This results in more realistically simulated soil moisture fields and better simulation of the dynamic range of the surface temperature when compared with the other model configurations.

Abstract

A highly modular and scale-consistent Terrestrial Systems Modeling Platform (TerrSysMP) is presented. The modeling platform consists of an atmospheric model (Consortium for Small-Scale Modeling; COSMO), a land surface model (the NCAR Community Land Model, version 3.5; CLM3.5), and a 3D variably saturated groundwater flow model (ParFlow). An external coupler (Ocean Atmosphere Sea Ice Soil, version 3.0; OASIS3) with multiple executable approaches is employed to couple the three independently developed component models, which intrinsically allows for a separation of temporal–spatial modeling scales and the coupling frequencies between the component models.

Idealized TerrSysMP simulations are presented, which focus on the interaction of key hydrologic processes, like runoff production (excess rainfall and saturation) at different hydrological modeling scales and the drawdown of the water table through groundwater pumping, with processes in the atmospheric boundary layer. The results show a strong linkage between integrated surface–groundwater dynamics, biogeophysical processes, and boundary layer evolution. The use of the mosaic approach for the hydrological component model (to resolve subgrid-scale topography) impacts simulated runoff production, soil moisture redistribution, and boundary layer evolution, which demonstrates the importance of hydrological modeling scales and thus the advantages of the coupling approach used in this study.

Real data simulations were carried out with TerrSysMP over the Rur catchment in Germany. The inclusion of the integrated surface–groundwater flow model results in systematic patterns in the root zone soil moisture, which influence exchange flux distributions and the ensuing atmospheric boundary layer development. In a first comparison to observations, the 3D model compared to the 1D model shows slightly improved predictions of surface fluxes and a strong sensitivity to the initial soil moisture content.

Abstract

An algorithm is presented to derive the downwelling solar surface irradiance from satellite measurements of the 0.63-μm reflectance, which explicitly accounts for variations in cloud optical depth and integrated water vapor. For validation, a long-term dataset of 40 356 pyranometer measurements and 1450 NOAA-14 Advanced Very High Resolution Radiometer (AVHRR) satellite scenes of the Netherlands is used. A mean overestimate of the satellite-retrieved irradiance by 7% is found, which is consistent with numerous other studies reporting positive biases of atmospheric transmissivities that are calculated by radiative transfer schemes in comparison with measurements. The bias can be explained by the calibration and measurement uncertainties of both the AVHRR and pyranometer. A strong solar zenith angle dependence of the bias is found when water clouds are assumed in the retrieval. Such a dependence is not observed for ice clouds. Currently, there is not enough information for a conclusive explanation of this behavior. Comparing individual pyranometer measurements at 30 stations within a region of about 150 km² averaged over 40 min, a large rmse of 86 W m⁻² is found. If the average of all of the stations for a satellite overpass is considered instead, a much better accuracy is achieved (rmse of 33 W m⁻²). For monthly averages of all of the stations, the rmse is further reduced to 12 W m⁻². Evidence is presented that suggests that a significant fraction of the rmse in the comparison originates from the variability of the irradiance field, which limits the representativeness of the reference ground-based pyranometer measurements for the satellite-retrieved values.

Abstract

Presented are quantitative estimates of specific attenuation and specific differential attenuation of 5-cm-wavelength radiation (C band) obtained by comparison with measurements at 10-cm wavelength (S band), which are much less affected by attenuation. The data originated from two almost-collocated radars in central Oklahoma. To avoid biases in estimates, the slopes with respect to range of differences in reflectivities and differential reflectivities are assumed to represent the specific attenuations. Observations on a day with no reports of hail on the ground and on a day with large hail are contrasted. A simple one-dimensional model of melting hail is used to qualify these observations. Examples of volumetric fields of the polarimetric variables obtained at the two wavelengths are presented to illustrate that much can be learned about size, orientation, and phase of hydrometeors over volumes that play a role in precipitation formation.