Cloud Size Distribution Research Articles

Reconciling Chord Length Distributions and Area Distributions for Fields of Fractal Cumulus Clouds

While the total cover of broken cloud fields can in principle be obtained from one-dimensional measurements, the cloud size distribution normally differs between two-dimensional (area) and one-dimensional retrieval (chord length) methods. In this study, we use output from high-resolution Large Eddy Simulations to generate a transfer function between the two. We retrieve chord lengths and areas for many clouds, and plot the one as a function of the other, and vice versa. We find that the cloud area distribution conditional on the chord length behaves like a gamma distribution with well-behaved parameters, with a mean μ=1.1L and a shape parameter β=L−0.645. Using this information, we are able to generate a transfer function that can adjust the chord length distribution so that it comes much closer to the cloud area distribution. Our transfer function improves the error in predicting the mean cloud size, and is performs without strong biases for smaller sample sizes. However, we find that the method is still has difficulties in accurately predicting the frequency of occurrence of the largest cloud sizes.

On the temperature influence on cavitation erosion in micro-channels

The effect of fuel temperature on the erosion intensity in micro-channels for different fuels like calibration fluid ISO 4113 and dodecane is investigated. A high-pressure closed loop test bench for fuels under elevated temperatures up to 160 °C is used to erode micro-channels made of Al 99.0 at a maximum pressure of 500 bar. The evaluation of the erosion intensity reveals a temperature of maximum erosion for each fuel. The temperature of maximum erosion correlates with the temperature dependence of the maximum collapse pressure obtained by the Rayleigh-Plesset equation. Infrared thermography shows a local temperature drop in the cavitation region of the channels. 2D CFD simulations with a thermodynamic cavitation model supplement the measurement results by an assessment of the temporal evolution of the temperature field. Similar to the measurements, the CFD results yield a local cooling by the thermodynamic effect and thus support the measurement results. At high fluid temperatures one crucial link between the progressing local temperature drop and the erosion reduction is pointed out: the change of the detached cloud size distribution with temperature.

Spatial distribution of cloud droplet size properties from Airborne Hyper-Angular Rainbow Polarimeter (AirHARP) measurements

Abstract. The global variability of clouds and their interactions with aerosol and radiation make them one of our largest sources of uncertainty related to global radiative forcing. The droplet size distribution (DSD) of clouds is an excellent proxy that connects cloud microphysical properties with radiative impacts on our climate. However, traditional radiometric instruments are information-limited in their DSD retrievals. Radiometric sensors can infer droplet effective radius directly but not the distribution width, which is an important parameter tied to the growth of a cloud field and to the onset of precipitation. DSD heterogeneity hidden inside large pixels, a lack of angular information, and the absence of polarization limit the amount of information these retrievals can provide. Next-generation instruments that can measure at narrow resolutions with multiple view angles on the same pixel, a broad swath, and sensitivity to the intensity and polarization of light are best situated to retrieve DSDs at the pixel level and over a wide spatial field. The Airborne Hyper-Angular Rainbow Polarimeter (HARP) is a wide-field-of-view imaging polarimeter instrument designed by the University of Maryland, Baltimore County (UMBC), for retrievals of cloud droplet size distribution properties over a wide swath, at narrow resolution, and at up to 60 unique, co-located view zenith angles in the 670 nm channel. The cloud droplet effective radius (CDR) and variance (CDV) of a unimodal gamma size distribution are inferred simultaneously by matching measurement to Mie polarized phase functions. For all targets with appropriate geometry, a retrieval is possible, and unprecedented spatial maps of CDR and CDV are made for cloud fields that stretch both across the swath and along the entirety of a flight observation. During the NASA Lake Michigan Ozone Study (LMOS) aircraft campaign in May–June 2017, the Airborne HARP (AirHARP) instrument observed a heterogeneous stratocumulus cloud field along the solar principal plane. Our retrievals from this dataset show that cloud DSD heterogeneity can occur at the 200 m scale, much smaller than the 1–2 km resolution of most spaceborne sensors. This heterogeneity at the sub-pixel level can create artificial broadening of the DSD in retrievals made at resolutions on the order of 0.5 to 1 km. This study, which uses the AirHARP instrument and its data as a proxy for upcoming HARP CubeSat and HARP2 spaceborne instruments, demonstrates the viability of the HARP concept to make cloud measurements at scales of individual clouds, with global coverage, and in a low-cost, compact CubeSat-sized payload.

Effect of admixed solid inertants on dispersibility of combustible dust clouds in a modified hartmann tube

Reduction of dust explosion hazards can be achieved by processing materials in less hazardous forms. Two principles are often employed to moderate explosion potential. One approach is to increase dust particle size so as to decrease its reactivity. The other involves altering the dust composition by admixture with solid inertants. An essential prerequisite for dust explosions is the formation of airborne dust clouds. Dust dispersibility (i.e., the ease of dispersion of a dust and the tendency of the particulate matter to remain airborne once a dust cloud has been formed) is therefore an important safety characteristic, certainly requiring research attention. This paper presents experimental results using a modified Hartmann tube to determine the effect of admixed solid inertants on particle size distribution of dust clouds. Nine types of combustible dust and two sizes of inert Al2O3 were tested repeatedly five times. Generally, admixing small amounts of micro-sized solid inertants caused relative improvement on the dispersibility of combustible dust, forming a better dispersed cloud, and sometimes resulted in increased ignition sensitivity. However, the addition of agglomerate nano inert particles caused a multilayer coating of combustible particles, causing an increase in the effective surface area at the particle contact points, and thereby increasing cohesion in the powder mixtures. Higher inter-particle forces in turn enhanced the combustible particle agglomeration during dispersion, leading to an increase in effective particle size distribution.

The Dependence of Shallow Cumulus Macrophysical Properties on Large‐Scale Meteorology as Observed in ASTER Imagery

AbstractThis study identifies meteorological variables that control the macrophysical properties of shallow cumulus cloud fields over the tropical ocean. We use 1,158 high‐resolution Advanced Spaceborn Thermal Emission and Reflection Radiometer (ASTER) images to derive properties of shallow cumuli, such as their size distribution, cloud top heights, fractal dimensions, and spatial organization, as well as cloud amount. The large‐scale meteorology is characterized by the lower‐tropospheric stability, subsidence rate, sea surface temperature, total column water vapor, wind speed, wind shear, and Bowen ratio. The surface wind speed emerges as the most powerful control factor. With increasing wind speed the cloud amount and cloud top heights show a robust increase accompanied by a marked shift in the cloud size distribution toward larger clouds with smoother shapes. These results lend observational support to the deepening response of a wind‐driven marine boundary layer as simulated by large‐eddy models. The other control factors cause smaller changes in the cloud field properties. We find a robust increase in cloud amount with increasing stability and decreasing sea surface temperature, respectively, which confirms a well‐known behavior of marine stratocumulus also for shallow cumulus clouds. Due to the high resolution of cloud images, we are able to study the lower end of the cloud size distribution and find a robust double power law behavior with a scale break at 590 m. We find a variation in the shape of the cloud size distribution with Bowen ratio, qualitatively consistent with modeling results and suggesting the Bowen ratio as a new potential control factor on shallow cumulus clouds.

Power-Law Scaling in the Internal Variability of Cumulus Cloud Size Distributions due to Subsampling and Spatial Organization

AbstractIn this study, the spatial structure of cumulus cloud populations is investigated using three-dimensional snapshots from large-domain LES experiments. The aim is to understand and quantify the internal variability in cloud size distributions due to subsampling effects and spatial organization. A set of idealized shallow cumulus cases is selected with varying degrees of spatial organization, including a slowly organizing marine precipitating case and five more quickly organizing diurnal cases over land. A subdomain analysis is applied, yielding cloud number distributions at sample sizes ranging from severely undersampled to nearly complete. A strong power-law scaling is found in the relation between cloud number variability and subdomain size, reflecting an inverse linear relation. Scaling subdomain size by cloud size yields a data collapse across time points and cases, highlighting the role played by cloud spacing in controlling the stochastic variability. Spatial organization acts on top of this baseline model by increasing the maximum cloud size and by enhancing the variability in the number of smallest clouds. This reflects that the smaller clouds start to live on top of larger-scale thermodynamic structures, such as cold pools, which favor or inhibit their formation. Compositing all continental cumulus cases suggests the existence of a prototype diurnal time dependence in the spatial organization. A simple stochastic expression for cloud number variability is proposed that is formulated in terms of two dimensionless groups, which allows objective estimation of the degree of spatial organization in simulated and observed cumulus cloud populations.

Investigating the Diurnal Evolution of the Cloud Size Distribution of Continental Cumulus Convection Using Multiday LES

Abstract The diurnal dependence of cumulus cloud size distributions over land is investigated by means of an ensemble of large-eddy simulations (LESs). A total of 146 days of transient continental shallow cumulus are selected and simulated, reflecting a low midday maximum of total cloud cover, weak synoptic forcing, and the absence of strong surface precipitation. The LESs are semi-idealized, forced by large-scale model output but using an interactive surface. This multitude of cases covers a large parameter space of environmental conditions, which is necessary for identifying any diurnal dependencies in cloud size distributions. A power-law exponential function is found to describe the shape of the cloud size distributions for these days well, with the exponential component capturing the departure from power-law scaling at the larger cloud sizes. To assess what controls the largest cloud size in the distribution, the correlation coefficients between the maximum cloud size and various candidate variables reflecting the boundary layer state are computed. The strongest correlation is found between total cloud cover and maximum cloud size. Studying the size density of the cloud area revealed that larger clouds contribute most to a larger total cloud cover, and not the smaller ones. Besides cloud cover, cloud-base and cloud-top height are also found to weakly correlate with the maximum cloud size, suggesting that the classic idea of deeper boundary layers accommodating larger convective thermals still holds for shallow cumulus. Sensitivity tests reveal that the results are only minimally affected by the representation of microphysics and the output resolution.

Origin of 1I/’Oumuamua. II. An Ejected Exo-Oort Cloud Object?

Abstract 1I/’Oumuamua is the first detected interstellar interloper. We test the hypothesis that it is representative of a background population of exo-Oort cloud objects ejected under the effect of post-main sequence mass loss and stellar encounters. We do this by comparing the cumulative number density of interstellar objects inferred from the detection of 1I/’Oumuamua to that expected from these two clearing processes. We consider the 0.08–8 M ⊙ mass range, take into account the dependencies with stellar mass, Galactocentric distance, and evolutionary state, and consider a wide range of size distributions for the ejected objects. Our conclusion is that 1I/’Oumuamua is likely not representative of this background population, even though there are large uncertainties in the masses and size distributions of the exo-Oort Clouds. We discuss whether the number density of free-floating, planetary-mass objects derived from gravitational microlensing surveys could be used as a discriminating measurement regarding 1I/’Oumuamua’s origin (given their potential common origin). We conclude that this is challenged by the mass limitation of the surveys and the resulting uncertainty of the mass distribution of the free floaters. The detection of interlopers may be one of the few observational constraints of the small end of this population, with the caveat that, as we conclude here and in Moro-Martín (2018), in the case of 1I/’Oumuamua, it might not be appropriate to assume this object is representative of an isotropic background population, which makes the derivation of a number density very challenging.

Thermodynamic Constraints on the Size Distributions of Tropical Clouds

AbstractTropical convective clouds evolve over a wide range of temporal and spatial scales, which makes them difficult to simulate numerically. Here we propose that cloud statistical properties can be derived within a simplified time‐independent coordinate system of cloud number n, saturated static energy h⋆, and cloud perimeter λ. Under the constraint that circulations around cloud edge compete with each other for total buoyant energy and air, we show that the product of cloud number and cloud perimeter nλ is invariant with λ and that cloud number follows a negative exponential with respect to cloud‐edge deviations of h⋆ with respect to the mean. Overall, the summed perimeter of all clouds scales as the square root of the atmospheric static stability. These theoretical results suggest that the complexity of cloud field structures can be viewed statistically as an emergent property of atmospheric bulk thermodynamics. Comparison with a detailed tropical cloud field simulation shows general agreement to within ≤13%. For the sake of developing hypotheses about cloud temporal evolution that are testable in high resolution simulations, the shapes of tropical cloud perimeter distributions are predicted to be invariant as climate warms, although with a modest increase in total cloud amount.

Measuring the detector-observed impact of optical blurring due to aerosols in a laboratory cloud chamber

Deleterious effects of the atmosphere on remotely acquired images includes absorption and scattering of light by aerosol particulates, which not only attenuates the signal but can potentially cause blurring due to forward-scattered light accepted by the imaging system. Proposed aerosol scattering models (e.g., Ishimaru) provide a method for simulating the contrast and spatial detail expected when imaging through atmospheres with significant aerosol optical depth. This work explores closure between modulation transfer functions (MTFs) obtained from directly measured images and MTFs calculated from theory using measured cloud properties. The closure experiments are performed in a laboratory cloud chamber in which cloud droplet number density and size distribution are directly measured. Images of a binary knife-edge target were taken with an optical detector on the other side of a water cloud generated through reduction of pressure in the humidified chamber. The key results of this closure experiment are: the theoretical expression for the aerosol MTF is likely overly simplistic and does not account for broad particle size distributions. The significance of optical blurring from light scattering by aerosol particles depends sensitively on the properties of both the particles and the imaging system.

Methods for Estimating 2D Cloud Size Distributions from 1D Observations

Abstract The two-dimensional (2D) size distribution of clouds in the horizontal plane plays a central role in the calculation of cloud cover, cloud radiative forcing, convective entrainment rates, and the likelihood of precipitation. Here, a simple method is proposed for calculating the area-weighted mean cloud size and for approximating the 2D size distribution from the 1D cloud-chord lengths measured by aircraft and vertically pointing lidar and radar. This simple method (which is exact for square clouds) compares favorably against the inverse Abel transform (which is exact for circular clouds) in the context of theoretical size distributions. Both methods also perform well when used to predict the size distribution of real clouds from a Landsat scene. When applied to a large number of Landsat scenes, the simple method is able to accurately estimate the mean cloud size. As a demonstration, the methods are applied to aircraft measurements of shallow cumuli during the Routine ARM Aerial Facility (AAF) Clouds with Low Optical Water Depths (CLOWD) Optical Radiative Observations (RACORO) campaign, which then allow for an estimate of the true area-weighted mean cloud size.

Analysis of albedo versus cloud fraction relationships in liquid water clouds using heuristic models and large eddy simulation

AbstractThe relationship between the albedo of a cloudy scene and cloud fraction fc is studied with the aid of heuristic models of stratocumulus and cumulus clouds. Existing work has shown that scene albedo increases monotonically with increasing cloud fraction but that the relationship varies from linear to superlinear. The reasons for these differences in functional dependence are traced to the relationship between cloud deepening and cloud widening. When clouds deepen with no significant increase in fc (e.g., in solid stratocumulus), the relationship between and fc is linear. When clouds widen as they deepen, as in cumulus cloud fields, the relationship is superlinear. A simple heuristic model of a cumulus cloud field with a power law size distribution shows that the superlinear ‐fc behavior is traced out either through random variation in cloud size distribution parameters or as the cloud field oscillates between a relative abundance of small clouds (steep slopes on a log‐log plot) and a relative abundance of large clouds (flat slopes). Oscillations of this kind manifest in large eddy simulation of trade wind cumulus where the slope and intercept of the power law fit to the cloud size distribution are highly correlated. Further analysis of the large eddy model‐generated cloud fields suggests that cumulus clouds grow larger and deeper as their underlying plumes aggregate; this is followed by breakup of large plumes and a tendency to smaller clouds. The cloud and thermal size distributions oscillate back and forth approximately in unison.

An introduction to light extinction spectrometry as a diagnostic for dust particle characterisation in dusty plasmas

In this article, a detailed introduction of the light extinction spectrometry (LES) diagnostics is given. LES allows the direct in situ measurement of the particle size distribution and absolute concentration of dust clouds levitating in plasmas. Using a relatively simple and compact experimental set-up, the dust cloud parameters can be recovered with a good accuracy making minimum assumptions on their physical properties. Special emphases are given to the inversion procedure of light extinction spectra and all the required particle shape, refractive index and light extinction models. The parameter range and the limitations of LES are discussed. Two measurements in low-pressure gas discharges are presented: (i) in a direct-current (DC) glow discharge in which nanoparticles are growing from the sputtering of a tungsten cathode and (ii) in an argon–silane radio-frequency discharge. They demonstrate the capabilities of the LES technique to characterise, in situ and in real-time, the growth dynamics of nanoparticles in the size range 5–100 nm and volume concentrations in the range from a few ppb to a few ppm.

Elongated dust clouds in a uniform DC positive column of low pressure gas discharge

Experimental investigations of the formation of elongated dust clouds and their influence on the plasma glow intensity of the uniform direct current (DC) positive column (PC) have been performed under microgravity conditions. For the axial stabilization of the dust cloud position a polarity switching DC gas discharge with a switching frequency of 250 Hz was used. During the experiment, a spontaneous division of one elongated dust cloud into two smaller steady state dust clouds has been observed. Quantitative data on the dust cloud shape, size and dust number density distribution were obtained. Axial and radial distributions of plasma emission within the 585.2 nm and 703.2 nm neon spectral lines were measured over the whole discharge volume. It has been found that both spectral line intensities at the dust cloud region grew 1.7 times with respect to the undisturbed positive column region; in this the 585.2 nm line intensity increased by 10% compared to the 703.2 nm line intensity. For a semi-quantitative explanation of the observed phenomena the Schottky approach based on the equation of diffusion was used. The model reasonably explains the observed glow enhancement as an increasing of the ionization rate in the discharge with dust cloud, which compensates ion-electron recombination on the dust grain surfaces. In this, the ionization rate increases due to the growing of the DC axial electric field, and the glow grows directly proportional to the electric field. It is shown that the fundamental condition of the radial stability of the dusty plasma cloud is equal to the ionization and recombination rates within the cloud volume that is possible only when the electron density is constant and the radial electric field is absent within the dust cloud.

Evaluation of tropical channel refinement using MPAS‐A aquaplanet simulations

AbstractClimate models with variable‐resolution grids offer a computationally less expensive way to provide more detailed information and increased accuracy by resolving processes that cannot be adequately represented by a coarser grid. This study uses the Model for Prediction Across Scales‐Atmosphere (MPAS‐A), consisting of a nonhydrostatic dynamical core and a subset of Weather Research and Forecasting (WRF) model physics, to investigate the potential benefits of using tropical channel refinement. The simulations are performed with an idealized aquaplanet configuration using 30 and 240 km global grid spacing, and two variable‐resolution grids spanning the same grid spacing range; one with a narrow (20°S–20°N) and one with a wide (30°S–30°N) tropical channel refinement. Increasing resolution in the tropics impacts both the tropical and extratropical circulation. Compared to the 30 km global grid, both refined channel simulations exhibit slightly stronger updrafts inside the Hadley cell resulting in more resolved precipitation. Using a wider tropical refinement leads to a closer correspondence with the global high‐resolution grid. While different grid spacings produce similar cloud size distributions that are consistent with observations, the dependence of precipitation rate on cloud size varies among simulations. The refined channel simulations show improved tropical and extratropical precipitation relative to the global coarse simulation. All simulations show a single precipitation peak centered on the equator. Although the results show that tropical refinement is an effective method for avoiding artifacts due to grid resolution sensitivities seen in earlier studies that only refined a portion of the tropics, some biases remain well inside of the refinement region.

THE ORIGIN OF DUST EXTINCTION CURVES WITH OR WITHOUT THE 2175 Å BUMP IN GALAXIES: THE CASE OF THE MAGELLANIC CLOUDS

The Large and Small Magellanic Clouds (LMC and SMC, respectively) are observed to have characteristic dust extinction curves that are quite different from those of the Galaxy (e.g., strength of the 2175 A bump). Although the dust composition and size distribution of the Magellanic Clouds (MCs) that can self-consistently explain their observed extinction curves have been already proposed, it remain unclear whether and how the required dust properties can be achieved in the formation histories of the MCs. We therefore investigate the time evolution of the dust properties of the MCs and thereby derive their extinction curves using one-zone chemical evolution models with formation and evolution of small and large silicate and carbonaceous dust grains and dusty winds associated with starburst events. We find that the observed SMC extinction curve without a conspicuous 2175 A bump can be reproduced well by our SMC model, if the small carbon grains can be selectively lost through the dust wind during the latest starburst about 0.2 Gyr ago. We also find that the LMC extinction curve with a weak 2175 A bump can be reproduced by our LMC model with less efficient removal of dust through dust wind. We discuss possible physical reasons for different dust wind efficiencies between silicate and graphite and among galaxies.

In-situ measurements of cloud-precipitation microphysics in the East Asian monsoon region since 1960

A large number of in-situ measurements of cloud-precipitation microphysical properties have been made since 1960, including measurements of particle size distribution, particle concentration, and liquid water content of clouds and rain. These measurements have contributed to considerable progress in understanding microphysical processes in clouds and precipitation and significant improvements in parameterizations of cloud microphysics in numerical models. This work reviews key findings regarding cloud-precipitation microphysics over China. The total number concentrations of various particles vary significantly, with certain characteristic spatial scales. The size distributions of cloud droplets in stratiform clouds can generally be fit with gamma distributions, but the fit parameters cover a wide range. Raindrop size distributions (RSDs) associated with stratiform clouds can be fit with either exponential or gamma distributions, while RSDs associated with convective or mixed stratiform-cumuliform clouds are best fit with gamma distributions. Concentrations of ice nuclei (IN) over China are higher than those observed over other regions, and increase exponentially as temperature decreases. The particle size distributions of ice crystals, snow crystals, and hailstones sampled at a variety of locations can be reliably approximated by using exponential distributions, while aerosol particle size distributions are best described as the sum of a modified gamma distribution and a Junge power-law distribution. These results are helpful for evaluating and improving the fidelity of physical processes and hydrometeor fields simulated by microphysical parameterizations. The comprehensive summary and analysis of previous work presented here also provide useful guidelines for the design of future observational programs.

A simple dynamical model of cumulus convection for data assimilation research

A simplified model for cumulus convection has been developed, with the aim of providing a computationally inexpensive, but physically plausible, environment for developing methods for convective-scale data assimilation. Key processes, including gravity waves, conditional instability and precipitation formation, are represented, and parameter values are chosen to reproduce the most important space and time scales of cumulus clouds. The model is shown to reproduce the classic life cycle of an isolated convective storm. When provided with a low amplitude noise source to trigger convection, the model produces a statistically steady state with cloud size and cloud spacing distributions similar to those found in radiative-convective equilibrium simulations using a cloud resolving model. Results are also shown for convection triggered by flow over an orgraphic obstacle, where depending on the wind speed two regimes are found with convection trapped over the mountain, or propagating downstream. The model features prognostic variables for wind and rain that can be used to compute synthetic observations for data assimilation experiments.

The Influence of Land Surface Heterogeneities on Cloud Size Development

Abstract This study analyzes the effects of land surface heterogeneities at various horizontal scales on the transition from shallow to deep convection and on the cloud size distribution. An idealized case of midlatitude summertime convection is simulated by means of large-eddy simulations coupled to an interactive land surface. The transition is accelerated over heterogeneous surfaces. The simulation with an intermediate patch size of 12.8 km exhibits the fastest transition with a transition time two-thirds that over a homogeneous surface. A similar timing is observed for the precipitation onset whereas the total accumulated rainfall tends to increase with patch size. The cloud size distribution can be approximated by a power law with a scale break. The exponent of the power law is independent of the heterogeneity scale, implying a similar cloud cover between the simulations. In contrast, the scale break varies with patch size. The size of the largest clouds does not scale with the boundary layer height, although their maximum size scales with the patch size. Finally, the idea that larger clouds grow faster, known from homogeneous surface conditions, is not fully valid over heterogeneous surfaces. These various aspects can be understood from the complex interplay between the characteristics of the triggered mesoscale circulations and a cloud development acting in response to the diurnal cycle in surface heating. The results also call for adequate representation of such effects in convective parameterizations.

Evaluation of Precipitating Hydrometeor Parameterizations in a Single-Moment Bulk Microphysics Scheme for Deep Convective Systems over the Tropical Central Pacific

Abstract Cloud microphysics of deep convective systems over the tropical central Pacific simulated by a cloud system–resolving model using satellite simulators are evaluated in terms of the joint histogram of cloud-top temperature and precipitation echo-top heights. A control experiment shows an underestimation of stratiform precipitation and a higher frequency of precipitating deep clouds with top heights higher than 12 km when compared with data from the Tropical Rainfall Measuring Mission. The comparison shows good agreement for horizontal distribution and statistical cloud size distributions of deep convective systems. Biases in the joint histogram are improved by changing cloud microphysics parameters of a single-moment bulk microphysics scheme. The effects of size distribution of precipitating hydrometeors are examined. Modification of the particle size distributions of rain, snow, and graupel size distributions based on observed relationships improves cloud precipitation statistics. This study implies that a single-moment bulk cloud microphysics scheme can be improved by employing comparison of satellite observations and diagnostic relationships.