7 Summary, Conclusions and Future Work same aspect ratio behave in a similar way. In comparison, absorption dominates the extinction in the case of liquid droplets. The absorption cross-section of liquid droplets is at least an order of magnitude larger than that of ice crystals.
The accuracy of the T-matrix method was checked by comparing it with the Mie code and the difference for extinction cross-section was smaller than 0.002 % at all frequencies ranging from 50–1000 GHz.
One limitation of the T-matrix method is that it does not converge for large size parameters and particles that are highly aspherical. More-over, it can be used only for rotationally symmetric particles. Although the Discrete Dipole Approximation (dda) method can yield the single scattering properties of more complex ice particle shapes and orienta-tions that are common in cirrus clouds, it is computationally demand-ing. Moreover, there are some known inaccuracies in the polarization related elements of the scattering matrix. However, a comparison of the T-matrix and the dda code is planned to be done as part of an estec Study (Buehler et al., 2003).
Chapter 4 of the thesis presented the new scattering version of the artsradiative transfer model. Theartsclear sky model was chosen as the platform for implementing the scattering version. There are two things that make the scattering implementation complicated. One is the dimensionality of the scattering problem. For example, unlike the gaseous absorption cross-section, the scattering cross-section is highly direction-dependent. Because the vector radiative transfer equation has to be solved, the extinction cross-section and the phase matrix are 4×4matrices and the absorption cross-section is a 4×1vector.
Moreover, all the cross-sections and the atmospheric fields vary in a three-dimensional space. Accordingly, higher order tensors are neces-sary to store most of the variables that are used in the scattering part.
The second factor that makes the scattering complicated is linked to the solution method. Because the scattering problem is solved using an iterative method, the number of iterations required for achieving convergence can vary. For very strongly scattering cases, the number of iterations required can be very large. Because of these two reasons, it was decided to confine the scattering problem to a finite region of the atmosphere called the cloud box. A lot of computer memory and
computational time could be saved by using this strategy. In addition, to improve the speed of the calculation, optimization algorithms like the sequential update are also implemented inarts. It is also possible to use the model for a 1-D scalar case.
The newartsmodel has the capability to simulate multiple scat-tering by cirrus clouds and it also takes into account the polarization due to scattering. The model is suitable for simulating up, down, and limb looking sensors. In order to take into account the horizontal in-homogeneity of cirrus clouds,artsis designed to solve the radiative transfer equation in a spherical three-dimensional atmosphere. The spherical three-dimensional atmospheric geometry is particularly im-portant in the case of limb-viewing sensors, although this thesis does not cover the limb-viewing aspects of arts. The results for limb are covered in detail in another PhD thesis, by Emde (2005). For the down-looking sensors it is important that at window frequencies, the effect of surface reflection is properly taken into account which is also taken care of inarts.
A detailed analysis which studied the impact of cirrus cloud char-acteristics like ice water content, particle size and shape, and cloud position on upwelling microwave radiation is presented in Chapter 5.
In order to assess the impact of each parameter separately, simula-tions were done by varying each of the parameters over a wide range that is physically possible for that parameter. The example frequen-cies selected for the simulations are 89.9, 150.9, 184.31, 186.31, and 190.31 GHz. These frequencies correspond to the center frequencies of the upper side band of channels 16, 17, 18, 19, and 20 of amsu-b sensor on-board noaa-15, 16, and 17 satellites. The results showed that the sensitivity to cloud properties depends on the water vapor emission absorption background and on the cloud microphysical and macrophysical properties.
The effect of cirrus clouds on upwelling microwave radiation is to de-crease the brightness temperature. This is because the ice particles in cirrus, which interact with microwave radiation mostly through scat-tering, scatter away the upwelling radiation from the instrument field of view. Among all the frequencies considered, 89 GHz is the least sen-sitive to cirrus clouds. At this frequency, the scattering cross-section is
7 Summary, Conclusions and Future Work almost equal to the absorption cross-section. Another frequency that is not very sensitive to cirrus clouds is 184 GHz, but due to a differ-ent reason. The reason is that this frequency which is highly sensitive to water vapor gets saturated above the cloud altitude. Only when the cloud altitude is above the sounding altitude, the effect of cirrus clouds is felt. The frequency that is the most sensitive to cirrus clouds is 190 GHz because the radiation passes almost entirely through the cloud and the scattering is much stronger than the absorption.
At all frequencies, the effect of increasing the ice water path in-creases the scattering effect, i.e., the brightness temperature depres-sion deepens. The effect of increasing the particle size increases the extinction due to scattering which also leads to deeper brightness tem-perature depression. When the particle shape deviates from spherical shape, the brightness temperature depression is deeper than compared to the spherical case. It was shown that a preferred orientation can give rise to a significant polarization signature compared to a ran-domly oriented case.
For frequencies that are sensitive to the surface and have absorp-tion of the same order as scattering, sometimes the presence of cir-rus clouds can lead to a brightness temperature enhancement. This happens when the underlying surface is radiometrically cold and the cloud is present at a very low altitude. In this case, the emission of the cloud against the radiometrically cold background leads to an increase in brightness temperature. For liquid clouds, this is always the case, i.e., brightness temperature enhancement. This is because they are highly absorbing and are located at very low altitudes. The simula-tions presented in Chapter 5 showed that the presence of liquid clouds always increased the brightness temperature at 89 GHz. For channels that are not surface sensitive, the extinction by liquid cloud droplets decreased the brightness temperature compared to the clear sky case.
But since liquid clouds are present at lower layers of the atmosphere, their presence is insignificant to 184 GHz, for instance. Chapter 5 also presented the effects of clouds for the same frequencies using the fields taken from one day of globalera-40 reanalysis data.
Finally, the validation of the radiative transfer model was done by comparing the simulated brightness temperature to collocated
amsu-bobserved brightness temperature. The atmospheric profiles of pres-sure, temperature, and humidity as well as the cloud properties like ice water content profiles and cloud altitudes were taken from the ukMet Office mesoscale model forecast fields. The results suggested that the radiative transfer model is able to simulate the brightness temperature depression seen in the observation, over regions that are associated with the presence of cirrus clouds. The brightness temper-ature enhancement seen in channel 16 due to the presence of liquid clouds is also correctly reproduced in the simulations. However, there are some discrepancies between the simulation and the observation.
One reason is that the mesoscale model does not give information on the particle size, shape or orientation. Therefore the assumption regarding these parameters can be different from the real scenario.
For particle size distribution, a parameterization proposed by McFar-quhar and Heymsfield (1997) was used. But this parameterization was developed based on observations of tropical cirrus. In the future, it is planned to use parameterizations that are developed specifically for midlatitude cirrus. There are no ways to determine the shape and ori-entation of the particle unless there are some in-situ measurements.
There are also uncertainties associated with the mesoscale model fore-cast fields. There is a time difference of 30 minutes between the model forecast time and the observation time. This can contribute to some of the discrepancies between the observation and the simulation. It is also possible that the model generated atmospheric fields and cloud characteristics are not exactly similar to the real scenario. Further-more, there are some processes that are not taken into account in the simulation, for example the precipitation. The inclusion of precipita-tion can help to make the comparison better for channels 16, 17 and 20. Another discrepancy is associated with the surface emissivity over land surfaces. It is desirable to use an emissivity model that can give the surface emissivity values over different kinds of land surfaces.
In the future, it is planned to validate theartsmodel by comparing it to amsu-b observations using in-situ measurements of cloud and atmospheric parameters. Once the validation process is complete, it is planned to use the model as a tool for the retrieval of cirrus cloud properties in the sub-mm wavelength range.
I would first of all like to thank my advisor Dr. Stefan Bühler for giving me an opportunity to do my Ph.D. in his group. He had been always a pillar of support to me since the very first day, both aca-demically and personally. I still remember the effort he took in the beginning for introducing me to radiative transfer theory and arts program. I also thank him for the trust and confidence he put in me when he asked me to manage the esa-rt study. Next, I would like to thank Claudia Emde with whom I worked together in the develop-ment of the scattering algorithm. In those initial days of developdevelop-ment, we shared a lot of exciting moments when we discovered new bugs or when we got interesting results. I am more than indebted to Oliver Lemke who helped me to solve a lot of problems in C++ programing.
He was always there to solve any computer related problems. I am also thankful to Christian Melsheimer who was always keen to answer my questions related to scattering and polarization.
I thank Prof. K. Künzi who agreed to review my thesis and for giving me useful comments and suggestions that improved the thesis.
I also thank Dr. Clemens Simmer for agreeing to review my thesis.
The artsdevelopment would not have been possible without the help of Patrick Eriksson from the Chalmers University of Gothenburg, who always was quick to respond to any questions. The help of Cory Davis from the University of Edinburgh who implemented the Monte Carlo algorithm inartsis very much appreciated. Thepyarts pack-age developed by Cory Davis was also used in some calculations and I am grateful to him for providing it.
Theamsudata and the mesoscale model data that was used for the validation of artswas provided by Stephen English of the uk Met Office. I thank him and his colleagues Una O’Keeffe and Amy
Do-195
Acknowledgements herty who were always willing to help me with details of the data. We are also engaged in an ongoing activity to compare the rttovscatt model toarts. I take this opportunity to thank Stephen English and the members of the Satellite Assimilation group of the ukMet Office for their hospitality during my two months stay in the Met Office.
Thanks to Michael Mishchenko and Steven Warren for making pub-licly available the T-matrix program and the refractive index program, respectively, which were extensively used in this thesis for the calcu-lation of single scattering properties.
This is also the best opportunity to thank all the former and current members of the SAT group for giving the best working atmosphere I can ever expect. They had made my stay in Bremen an unforgettable experience. I thank especially my colleagues who helped me to proof-read my thesis. Thanks to Stefan Bühler, Claudia Emde and Viju Oommen John who spent a lot of time to read all chapters of my thesis.
Thanks also to Nathalie Courcoux, Mashrab Kuvatov, and Loknath Lamsal who helped in the editing of some chapters.
Thanks to Prof. Bleck Neuhaus and the PIP program which helped international students like me to assimilate easily into the life inside the institute.
My life in Bremen would not have been so much fun without my friends Viju Oommen John, Sandip Dhomse, Loknath Lamsal, Ni-nad Sheode, Semeena Valiayaveettil, Beena Balan, and Nizy Mathew.
Special thanks to Viju Oommen John who was always there during my moments of agony and ecstasy. I also thank Stefanie Bühler who had always an ear to listen to my problems. I enjoyed the occassional lunches and tea breaks we had together.
This acknowledgement would be incomplete if I don’t thank my parents and sister for their constant encouragement and support. Mere words are not enough to explain how they were a constant source of motivation for me.
This work was funded by the German Federal Ministry of Education and Research (bmbf), within the dlr project smiles, grant 50 ee 9815, and within theafo2000 projectuthmos, grant 07ATC04. It is also a contribution to the cost Action 723 ‘Data Exploitation and Modeling for the Upper Troposphere and Lower Stratosphere’.
Anderson, G. P., Clough, S. A., Kneizys, F. X., Chetwynd, J. H.
and Shettle, E. P., 1986: AFGL atmospheric constituent profiles (0–120 km). Tech. Rep. TR-86-0110, AFGL.
Bauer, P., 2002:Microwave radiative transfer modeling in clouds and precipitation Part 1: Model Description. Tech. Rep., NWP SAF Document No. NWPSAF-EC-TR-005.
Bennartz, R. and Bauer, P., 2003: Sensitivity of microwave radiances at 85-183 GHz to precipitating ice particles.Radio Science, 38, 4, 8075.
Bohren, C. F. and Huffman, D. R., 1983: Absorption and Scattering of Light by Small Particles. Wiley Interscience.
Brussard, G. and Watson, P. A., 1995: Atmospheric modelling and milli-metre wave propagation. Chapman and Hall.
Buehler, S., Emde, C., Schulz, J., Eriksson, P., English, S. and Davis, C., 2003:Development of a radiative transfer model for frequencies between 200 and 1000 GHz. A proposal in response to ESA-ESTEC invitation to tender AO/1-4320/03/NL/FF.
Buehler, S. A., Eriksson, P., Kuhn, T., von Engeln, A. and Verdes, C., 2005: ARTS, the Atmospheric Radiative Transfer Simulator.J.
Quant. Spectrosc. and Radiat. Transfer,91, 1, 65–93.
Buehler, S. A., Kuvatov, M., John, V. O., Leiterer, U. and Dier, H., 2004: Comparison of Microwave Satellite Humidity Data and Ra-diosonde Profiles: A Case Study.J. Geophys. Res.,109.
Burns, B. A., Wu, X. and Diak, G. R., 1997: Effects of precipitation and cloud ice on brightness temperatures in AMSU moisture chan-nels.IEEE Trans. Geosci. Remote Sensing,35, 6, 1429–1437.
Chandrasekhar, S., 1960:Radiative Transfer. Dover Publications.
Chepfer, H., Brogniez, G., Goloub, P., Bréon, F. M. and Flamant,
197
BIBLIOGRAPHY P. H., 1999: Observations of horizontally oriented ice crystals in cirrus clouds with POLDER-1/ADEOS-1.J. Quant. Spectrosc. and Radiat. Transfer,63, 521–543.
Comstock, J. M. and Jakob, C., 2004: Evaluation of tropical cirrus cloud properties derived from ECMWF model output and ground based measurements over Nauru Island. Geophys. Res. Lett.,31.
Czekala, H., 1999:Microwave radiative transfer calculations with mul-tiple scattering by nonspherical hydrometeors. Ph.D. thesis, Rheinis-che Friedrich-Wilhems-Universität Bonn.
Czekala, H. and Simmer, C., 1998: Microwave radiative transfer with nonspherical precipitating hydrometeors. J. Quant. Spectrosc. and Radiat. Transfer,60, 3, 365–374.
Davis, C., Emde, C. and Harwood, R., 2005: A 3D Polarized Reversed Monte Carlo Radiative Transfer Model for mm and sub-mm Pas-sive Remote Sensing in Cloudy Atmospheres.IEEE Trans. Geosci.
Remote Sensing,43, 6, 1096–1101.
Deeter, M. N. and Evans, K. F., 1998: A hybrid Eddington-single scattering radiative transfer model for computing radiances from thermally emitting atmospheres. J. Quant. Spectrosc. and Radiat.
Transfer,60, 4, 635–648.
Deschamps, P. Y., Breon, F. M., Leroy, M., Podaire, A., Bricaud, A., Buriez, J. C. and Seze, G., 1994: The POLDER Mission: Instru-ment characteristics and scientific objectives.IEEE Trans. Geosci.
Remote Sensing,2, 3, 598–615.
Diak, G. R., Kim, D., Whipple, M. S. and Wu, X., 1992: Preparing for the AMSU.Bull. Amer. Meteorol. Soc.,73, 1971–1984.
Donovan, D. P., 2003: Ice-cloud effective particle size parameteriza-tion based on combined lidar, radar reflectivity, and mean Doppler velocity measurements. J. Geophys. Res.,108, 18.
Dowling, D. R. and Radke, L. F., 1990: A summary of the physical properties of cirrus clouds.J. Appl. Met., 29, 970–978.
Draine, B. T., 2000: The Discrete Dipole Approximation for Light Scattering by Irregular Targets. In Light Scattering by Nonspheri-cal Particles: Theory, Measurements and Applications, Mishchenko, M. I., Hovenier, J. W. and Travis, L. D., eds., Academic Press, 131–
144.
Emde, C., 2005: A polarized discrete ordinate scattering model for simulations of limb and nadir long-wave measurements of cloudy atmospheres. Ph.D. thesis, University of Bremen.
Emde, C., Buehler, S. A., Davis, C., Eriksson, P., Sreerekha, T. R.
and Teichmann, C., 2004: A Polarized Discrete Ordinate Scattering Model for Simulations of Limb and Nadir Longwave Measurements in 1D/3D Spherical Atmospheres.J. Geophys. Res.,109, D24.
Emde, C. and Sreerekha, T. R., 2004: WP1.2 Model Review. Tech.
Rep., ESTEC. Development of a RT model for frequencies between 200 and 1000 GHz, Progress Report 1.
English, S. J. and Hewison, T. J., 1998: A fast generic millimetre-wave emissivity model. InProc. SPIE, 3503.
Eriksson, P., Buehler, S. A., Emde, C., Sreerekha, T. R., Melsheimer, C. and Lemke, O., 2003: ARTS-1-1 User Guide. University of Bremen. 308 pages, regularly updated versions available at www.sat.uni-bremen.de/arts/.
Evans, K. F., 1998: The spherical harmonics discrete ordinate method for three-dimensional atmospheric radiative transfer.J. Atmos. Sci., 55, 429–466.
Evans, K. F., Evans, A. H., Nolt, I. G. and Marshall, B. T., 1999: The Prospect for remote sensing of cirrus clouds with a submillimeter-wave spectrometer.J. Appl. Met.,38, 514–525.
Evans, K. F. and Stephens, G. L., 1991: A new polarized atmospheric radiative transfer model.J. Quant. Spectrosc. and Radiat. Transfer, 46, 5, 412–423.
Evans, K. F. and Stephens, G. L., 1995a: Microwave Radiative Trans-fer through Clouds Composed of Realistically Shaped Ice Crystals.
Part I: Single Scattering Properties.J. Atmos. Sci.,52, 11, 2041–
2057.
Evans, K. F. and Stephens, G. L., 1995b: Microwave Radiative Trans-fer through Clouds Composed of Realistically Shaped Ice Crystals.
Part II: Remote Sensing of Ice Clouds.J. Atmos. Sci.,52, 11, 2058–
2072.
Evans, K. F., Walter, S. J., Heymsfield, A. J. and Deeter, M. N., 1998:
Modeling of submillimeter passive remote sensing of cirrus clouds.
J. Appl. Met.,37, 184–205.