Submitted by Vinicius Machado Martinez null (vinicius@ipmet.unesp.br) on 2016-03-11T20:20:53Z No. of bitstreams: 1 defesa.pdf: 4478778 bytes, checksum: 15d0e211ad6c8ad44eabe8079cf63fa2 (MD5) / Approved for entry into archive by Juliano Benedito Ferreira (julianoferreira@reitoria.unesp.br) on 2016-03-15T17:11:38Z (GMT) No. of bitstreams: 1 martinez_vm_me_soro.pdf: 4478778 bytes, checksum: 15d0e211ad6c8ad44eabe8079cf63fa2 (MD5) / Made available in DSpace on 2016-03-15T17:11:38Z (GMT). No. of bitstreams: 1 martinez_vm_me_soro.pdf: 4478778 bytes, checksum: 15d0e211ad6c8ad44eabe8079cf63fa2 (MD5) Previous issue date: 2016-01-25 / O campo de refletividade de ecos observados por radares meteorológicos está sujeito a interferências de fenômenos físicos da atmosfera que podem resultar em interpretações não realísticas do fenômeno observado. Buscando ajustar o campo de refletividade de ecos detectados simultaneamente por dois radares meteorológicos ao longo da distância, este estudo desenvolveu um sistema baseado em regras fuzzy (SBRF), do tipo Takagi-Sugeno de primeira ordem, que combina as variáveis distância (km) e refletividade (dBZ) para expressar a refletividade ajustada de alvos mais distantes de um radar em relação a outro radar mais próximo, de modo que os efeitos das interferências nas medidas dos radares possam ser minimizados. Os dados utilizados são oriundos de dois radares meteorológicos do IPMet/UNESP, localizados nos municípios de Bauru (22,3583° S; 49,0278° W) e Presidente Prudente (22.175°1 S; 51.3743° W), no Brasil, no período de um ano de dados (2010) do produto CAPPI, faixa de 3.5km de altitude. A saída do sistema é nomeada refletividade fuzzy (RF) e é obtida através de um conjunto de nove curvas de regressão linear, cujos coeficientes angulares e lineares foram estimados pelo método dos mínimos quadrados. Dois parâmetros foram utilizados para análise das curvas obtidas: o coeficiente de correlação de Pearson e o coeficiente de determinação. O sistema foi aplicado a 20.514 dados referentes a 18 pixels distribuídos sobre uma faixa de comum cobertura dos radares. O modelo foi avaliado através das estatísticas de erro médio (BIAS), erro quadrático médio (MSE) e pelo teste de Kolmogorov-Smirnov. Os resultados obtidos demonstram a capacidade do sistema em aproximar o campo de refletividade de dois radares que operam sobre uma área de comum cobertura, devendo constituir-se como uma ferramenta alternativa de interpretação no monitoramento de chuvas, nos processos de modelagem ambiental e em sugestões futuras de estimativa de chuva por radar. / The reflectivity field of echoes observed by weather radars is subjected to the interference of physical phenomena of the atmosphere that can result in unrealistic interpretations on its characteristics. Aiming to adjust the reflectivity field of echoes simultaneously detected by two weather radars over a distance, this study developed a Takagi-Sugeno fuzzy rule-based system, whose input variables are the distance(km) and the reflectivity(dBZ) of the echoes, in order to obtain the adjusted reflectivity of echoes more distant from a radar in relation to another closer radar that is considered to be less susceptible to interference and, therefore, more realistic. The data used are from two weather radars of IPMet/UNESP, located in the municipalities of Bauru (22.3583° S; 49.0278° W) and Presidente Prudente (22.1751° S; 51.3743° W), Brazil, which collected data of the CAPPI product in the period of one year (2010), with a sampling altitude range of 3.5 km. The system output is named “Fuzzy Reflectivity” (FR), obtained through the fuzzy approach of a set of nine linear regression curves, whose angular and linear coefficients were estimated by the method of least squares. Two parameters were used for the analysis of the curves obtained: the Pearson correlation coefficient and the coefficient of determination. The system was applied to 20,514 data related to 18 pixels spread over a range of common coverage of the radars. We evaluated the system performance by means of statistical parameters: average error (bias), mean square error (MSE) and Kolmogorov-Smirnov test. The results obtained demonstrate the system ability to refine the modeling of the issue in question in relation to the traditional statistical approaches, properly adjusting the reflectivity field of the echoes observed by two radars that operate over an area of common coverage and serving as an alternative interpretation tool in the monitoring of rains, in processes of environmental modeling and in future suggestions of rain estimate by radar.

Au cours de l’automne 2012, un réseau d’observation très complémentaire a été déployé dans la région des Cévennes pour la période d’observation spéciale (SOP) du projet HyMeX. Ce réseau d’observation a été spécifiquement élaboré afin d’étudier la structure et l’hétérogénéité des précipitations et en particulier, l’impact du relief sur cette structure. Dans un premier temps, l’analyse de la distribution des gouttes de pluie (DSD) au sol et le long d’un profil vertical à partir des observations nous permet de décrire précisément la structure des précipitations le long d’un gradient topographique. Afin de comprendre l’influence du relief sur cette structure, nous nous concentrons sur les processus microphysique associés à la structure des précipitations. Pour ce faire, nous définissons trois régimes de pluie et étudions l’évolution verticale de la DSD le long du gradient topographique. Les variations en nombre ou en taille dans la DSD peuvent être associées à différents processus microphysiques ou dynamiques. Pour finir, nous estimons la capacité d’un modèle paramétré de l’atmosphère tel que WRF à représenter la structure des précipitations et les processus associés dans une zone de montagne. / During Fall 2012, a complementary observational network has been deployed in the Cévennes region (South of France) for the Special Observation Period (SOP) of the HyMeX project. This network has been specifically designed to study the structure and heterogeneity of precipitations and, in particulat, the impact of orography on this structure. Firstly, the analysis of the Drop Size Distribution (DSD) at the ground et along a vertical profile from ground observations allow us to describe precisely the rainfall structure along a topographical gradient. In order to understand the influence of a relief on this structure, we focus our study on the microphysical process associated with the structure of precipitations. To do so, we define three rainfall regime et study the vertical evolution of the DSD along the topographical gradient. Variations in number and size of the DSD can be associated with different microphysical or dynamical process. Finally, we estimate the capacity of a bulk atmospheric model such as WRF to represent the rainfall structure and associated mechanisms above a mountainous area.

La prévision quantitative des précipitations à l’aide des modèles météorologiques reste encore un grand défi posé à la communauté des sciences atmosphériques. En effet, deux problèmes majeurs sont généralement identifiés pour la prévision opérationnelle des précipitations et du climat : les interactions des systèmes précipitants avec le relief et avec la pollution. Cette thèse contribue à l’amélioration des prévisions de pluies. La stratégie adoptée est d’étudier des évènements précipitants en zones montagneuses en décrivant au mieux les interactions aérosol-nuage-précipitation à l’aide du modèle à microphysique mixte détaillée : DEtailed SCAvenging Model (DESCAM, Flossmann et Wobrock (2010)). Ce modèle utilise cinq distributions pour représenter les particules d’aérosol résiduelles et interstitielles ainsi que les gouttes et cristaux de glace. Le modèle a directement été comparé aux observations réalisées au cours de la campagne expérimentale COPS (Convective and Orographically induced Precipitation Study), qui a eu lieu pendant l’été 2007 à la frontière franco-allemande. En particulier, les simulations des pluies ont été comparées avec des observations de différents radars afin d’évaluer les performances du modèle mais aussi d’aider à l’interprétation des réflectivités de la bande brillante. La sensibilité par rapport à la pollution particulaire a été étudiée pour les propriétés des nuages et des précipitations. Pour les cas étudiés, plus le nombre des particules d’aérosol présentes dans l’atmosphère est important et plus leur solubilité est élevée, plus les précipitations au sol sont faibles. Ces comportements globaux peuvent toutefois être localement différents. Il existe donc des interactions plus complexes entre les particules d’aérosol, les nuages et les précipitations qui doivent être encore plus approfondies. / The quantitative precipitation forecast is still an important challenge for the atmospheric community. Indeed, two main problems are generally identified for weather and climate models : the interactions of the cloud systems with the topography and with pollution. This work contributes towards the improvement of the precipitation forecasts. The strategy used was to study the convective system over an area with a complex topography using the detailed microphysics scheme DEtailed SCAvenging Model (DESCAM, Flossmann and Wobrock (2010)) to better describe the aerosol-cloud-precipitation interactions. This microphysical scheme follows the evolution of the aerosol particle, drop and ice crystal distributions. Aerosol mass in drops and ice crystals is predicted by two distributions functions in order to close the aerosol budget. The model simulation results are compared with observations from COPS campaign (Convective and Orographically induced Precipitation Study), which took place at the French-German boarder during summer 2007. Rain simulations were compared with available radar data to evaluate the model’s performances and help the interpretation of the radar reflectivity in the bright band level. Sensitivity with respect to the particulate pollution was studied for in-cloud and precipitation properties. For the cases studied, the higher the aerosol particle number in the atmosphere or the higher the solubility of the aerosol particles, the weakest are the precipitation at the ground. These global behaviours of precipitation on the ground could be locally different. Consequently, the aerosol-cloud-precipitation interactions are complex and more in-depth studies are necessary.

Diese (Doktor)arbeit beschäftigt sich mit Radar-Polarimetrie, insbesondere mit der Untersuchung der Eigenschaften von polarimetrischen Variablen, die potenziellen Nutzen für die Radar-Meteorologie haben. Für den Einsatz in Dual-Polarisations-Radargeräten wird der Polarisationsgrad analysiert. Diese Variable wird in künftigen operationellen Radargeräten verfügbar sein. Der Polarisationsgrad hängt vom transmittierten Polarisationszustand und in weiterer Folge auch vom Betriebsmodus des Radargeräts ab. Der Hauptbetriebsmodus von operationellen Radargeräten sendet und empfängt gleichzeitig sowohl die horizontale als auch die vertikale Komponente. Der sekundäre Betriebsmodus sendet und empfängt simultan die horizontal polarisierte Komponente. In dieser Arbeit werden beide Polarisationsgrade untersucht. Da operationelle Systeme derzeit auf den Dual-Polarisationsmodus aufgerüstet werden, sollte künftig die Anwendungsmöglichkeiten von vollpolarimetrischen Wetterradarsystemen untersucht werden. Aus allen Variablen, die in diesem Betriebsmodus zur Verfügung stehen, wurde die Entropie (des gemessen Objektes) ausgewählt und wegen seiner engen Beziehung zum Polarisationsgrad näher untersucht. / The present doctoral thesis deals with radar polarimetry, namely with the investigation of properties of polarimetric variables potentially useful in radar meteorology. For use with dual-polarization radars, the degree of polarization is analyzed. This variable is available to planned operational radars. The degree of polarization is dependent on transmit polarization state and, consequently, it is dependent on the radar system operating mode. The primary operating mode of operational radars consists in simultaneous transmission and simultaneous receive of both horizontal and vertical components. The secondary operating mode consists of horizontal transmission and simultaneous receive. Both degrees of polarization are investigated in this thesis. Also, as operational systems are being updated to dual-polarization, research should start investigating the capabilities of fully polarimetric weather radar systems. Among the numerous variables available from this operating mode, the target entropy was chosen for investigation, also because of its close relation to the degree of polarization

The thesis is divided in two parts. The first part deals with the areal distribution of short-term convective rainfalls with regard to the influence of altitude. Precipitation estimates based on combination of rain gauge and radar data are used for this purpose. Statistical tests proved that the areal distribution of hourly convective rainfalls does not depend on altitude. Besides data containing precipitation events only, all measured data were statistically analysed regardless of the fact whether precipitation occurred or not. In this case it was found out that the relationship between hourly rainfall totals and altitude depends on the considered threshold of rainfall totals. When all data were considered, i.e. a threshold value was set to zero, an increase of rainfall totals well correlated with altitude. The dependence slowly disappeared with an increasing threshold. The areal distribution of 6 hour rainfall totals proved higher values in the area of south Bohemia. The most frequent synoptic patterns were northwest cyclonic situations (NWC) and cyclone over the Central Europe (C). The second part of the thesis is focused on satellite data exploitation, as measured by meteorological satellite Meteosat Second Generation, for convective precipitation estimates. The Convective Rainfall Rate (CRR) algorithm,...

Die Datenerfassung für Wetterprognosen basiert bis heute auf konventionellen Radarsystemen, die mit einer verhältnismäßig hohen Leistung arbeiten und für große Reichweiten ausgelegt sind. Da jedoch Wetterphänomene primär in Bodennähe auftreten und deren ausschlaggebenden Charakteristika ebendort zu detektieren sind, bringt dieses einige Nachteile mit sich. Hierzu zählen z.B. Einschränkungen bezüglich der räumlichen Auflösung und der Aktualisierungsrate, die stark eingeschränkten Möglichkeiten der flächendeckenden Erfassung bodennaher Effekte, aber auch die nicht voll polarimetrischen Detektionsmöglichkeiten bestehender Systeme. Die vorliegende Arbeit stellt den Ansatz des 'Digital Beamforming Weather Radar' (DB-WR) vor, welcher die beschriebenen Nachteile maßgeblich reduziert bzw. vermeidet. Die Systemarchitektur basiert hierbei auf engmaschigen Netzwerken von Phased-Array Radargeräten mit signifikant geringeren Reichweiten und Sendeleistungen. Grundlage hierfür bilden polarimetrische Sende-Empfangsmodule ('Software-Defined Radars'), welche die Realisierung der neuartigen 'Arbiträren Polarimetrischen Multiparameter Intrapulsmodulation' (APMIM), einem Verfahren welches beliebige Modulationen innerhalb des Sendepulses zulässt, ermöglichen. Der Fokus richtet sich diesbezüglich auf die Umsetzung eines breitbandigen Stand-Alone Experimentalsystems für diese neuartige Wetterradartechnologie, mit dem das Systemkonzept des DBWR getestet und die Möglichkeiten der APMIM in Kombination mit einer multiplen Empfangssignalauswertung evaluiert werden können. Darüber hinaus werden die Möglichkeiten dieses Experimentalsystems veranschaulicht und die Funktionalitäten in entsprechenden Messungen verifiziert. / Data acquisition for weather forecasts is still based on conventional radar systems, which operate at a relatively high power and are designed for long ranges. However, since weather phenomena primarily occur near the ground and their decisive characteristics have to be detected there, this brings with it a number of disadvantages. These include, for example, limitations with respect to spatial resolution and update rate, the severely restricted possibilities of area-wide detection of near-ground effects, but also the not fully polarimetric detection capabilities of existing systems. This dissertation presents the Digital Beamforming Weather Radar (DBWR) approach, which significantly reduces or avoids the described drawbacks. The system architecture is based on close-meshed networks of phased-array radars with significantly lower ranges and transmission powers. The basis for this is formed by polarimetric transmit-receive modules ('Software-Defined Radars'), which enable the realization of the novel 'Arbitrary Polarimetric Multiparameter Intrapulse Modulation' (APMIM), a method which allows arbitrary modulations within the transmit pulse. In this respect, the focus is on the implementation of a broadband stand-alone experimental system for this novel weather radar technology, with which the system concept of the DBWR can be tested and the possibilities of the APMIM in combination with a multiple received signal evaluation can be evaluated. Furthermore, the capabilities of this experimental system are illustrated and the functionalities are verified in corresponding measurements.

In this study a calibration concept for a multi-polarimetric weather radar is developed. Several common calibration techniques are analysed, but many are insufficient due to the non-reciprocal behaviour of the employed radar. Hence, an electronic calibration device was developed, which was designed for fast polarization determination of any polarization (including elliptical ones). The non-reciprocal behaviour was overcome by splitting receive and transmit calibration, which virtually uses the radar as a communication system. Beside the calibration a new and exible signal processing system was implemented on that radar which allows interleaved measurements using several polarimetric modes. This capability was used to analyse the STAR (hybrid basis with linear 45° transmit and horizontal/vertical receive) mode and the alternating H/V mode with respect to depolarization. Although it is known that depolarization causes errors in STAR mode, it is used in most commercial weather radars. / In dieser Arbeit wird ein Kalibrierkonzept für ein Multipolarisation-Radar entwickelt. Dazu wurden verschiedene gebräuchliche Techniken untersucht. Dabei stellte sich heraus, dass dieses Verfahren für das untersuchte nichtreziproke Radar unzureichend sind. Deshalb wurde ein elektronisches Kalibriergerät entwickelt, welches speziell der schnellen Messung von beliebigen Polarisationen - einschließlich Elliptischer - dient. Das nichtreziproke Verhalten wurde durch die Aufteilung in eine Sende- und eine Empfangskalibrierung umgangen, wodurch das Radar praktisch als Kommunikationssystem verwendet wird. Des Weiteren wurde eine neue, fexible Signalverarbeitung an dem Radar entwickelt, welches gemischte Messungen mit mehreren Polarisationsmoden erlaubt. Diese neuartige Möglichkeit wurde benutzt um den STAR-Modus, welches eine hybride Polarisationsbasis (linear 45° senden, horizontal/vertikal empfangen) benutzt, mit dem alternierende H/V-Modus zu vergleichen. Dabei wurde speziell das Verhalten des STAR-Modus im Hinblick auf Depolarisation untersucht, da dies bekanntermaßen zu Fehlern in den Messgrößenführen kann. Dies ist von besonderem Interesse, da der STAR-Modus in den meisten kommerziellen Wetterradarsystemen eingesetzt wird.

Flooding is one of the deadliest natural hazards in the world. Forecasting floods in advance can significantly reduce the socio-economic impacts. An accurate and reliable flood forecasting system is heavily dependent on the input precipitation data. Real-time, spatially, and temporally continuous Radar Quantitative Precipitation Estimates (QPEs) is useful precipitation information source. This research aims to investigate the efficacy of American and Canadian weather radar QPEs on hydrological model calibration and validation for flood forecasting in urban and semi-urban watersheds in Canada. A comprehensive review was conducted on the weather Radar network and its’ hydrological applications, challenges, and potential future research in Canada. First, radar QPEs were evaluated to verify the reliability and accuracy as precipitation input for hydrometeorological models. Then, the radar-gauge merging techniques were assessed to select the best method for urban flood forecasting applications. After that, merged Radar QPEs were used as precipitation input for the hydrological models to assess the impact of radar QPEs on hydrological model calibration and validation. Finally, a framework was developed by integrating hydrological and hydraulic models to produce flood forecasts and inundation maps in urbanized watersheds. Results indicated that dual-polarized radar QPEs could be effectively used as a source of precipitation input to hydrological models. The radar-gauge merging enhances both the accuracy and reliability of Radar QPEs, and therefore, the accuracy of streamflow simulation is also improved. Since flood forecasting agencies usually use hydrological models calibrated and validated using gauge data, it is recommended to use bias-corrected Radar QPEs to run existing hydrological models to simulate streamflow to produce flood extent maps. The hydrological and hydraulic models could be integrated into one framework using bias-corrected Radar QPEs to develop a successful flood forecasting system. / Thesis / Doctor of Science (PhD) / Floods are common and increasing deadly natural hazards in the world. Predicting floods in advance using Flood Early Warning System (FEWS) can facilitate flood mitigation. Radar Quantitative Precipitation Estimates (QPEs) can provide real-time, spatially, and temporally continuous precipitation data. This research focuses on bias-correcting and evaluating radar QPEs for hydrologic forecasting. The corrected QPE are applied into a framework connecting hydrological and hydraulic models for operational flood forecasting in urban watersheds in Canada. The key contributions include: (1) Dual-polarized radar QPEs is a useful precipitation input to calibrate, validate and run hydrological models; (2) Radar-gauge merging enhance accuracy and reliability of radar QPEs; (3) Floods could be more accurately predicted by integrating hydrological and hydraulic models in one framework using bias-corrected Radar QPEs; and (4) Gauge-calibrated hydrological models can be run effectively using the bias-corrected radar QPEs. This research will benefit future applications of real-time radar QPEs in operational FEWS.

<p> It has been hypothesized that some measurable properties of $Z_{DR}$ arcs in supercells may change in the minutes prior to tornadogenesis and tornadogenesis failure, and that $Z_{DR}$ arc area will change with SRH and can be used as a real-time proxy to estimate SRH. Output form the Cloud Model 1 (CM1) along with a polarimetric emulator is used to simulate $Z_{DR}$ arcs in 9 tornadic and 9 non-tornadic supercells. A random forest algorithm is used to automatically identify the $Z_{DR}$ arcs. Finally the inflow sector SRH is calculated at times when $Z_{DR}$ arcs are identified. To analyze the change in intensity and area a comparison between the average $Z_{DR}$ value inside and outside of the arc, as well as the spatial size of the arc and storm was done. Model calculated SRH is then compared to these metrics.</p> <p> </p> <p> It has also been observed that hail fallout complicates the automatic identification of $Z_{DR}$ arcs. In this study, three experiments are run where the simulated $Z_{DR}$ arcs are produced. One using all categories of hydrometeors, one where wet growth and melting of hail is excluded, and one excluding the contribution to $Z_{DR}$ from the hail hydrometeor category. The same analysis as above is repeated for all three experiments. Finally observed $Z_{DR}$ arcs are analyzed to see if these results are applicable to the real world. </p>

Recent advances in cloud-based big-data technologies now makes data driven solutions feasible for increasing numbers of scientific computing applications. One such data driven solution approach is machine learning where patterns in large data sets are brought to the surface by finding complex mathematical relationships within the data. Nowcasting or short-term prediction of rainfall in a given region is an important problem in meteorology. In this thesis we explore the nowcasting problem through a data driven approach by formulating it as a machine learning problem. State-of-the-art nowcasting systems today are based on numerical models which describe the physical processes leading to precipitation or on weather radar extrapolation techniques that predict future radar precipitation maps by advecting from a sequence of past maps. These techniques, while they can perform well over very short prediction horizons (minutes) or very long horizons (hours to days), tend not to perform well over medium horizons (1-2 hours) due to lack of input data at the necessary spatial and temporal scales for the numerical prediction methods or due to the inability of radar extrapolation methods to predict storm growth and decay. Given that water must first concentrate in the atmosphere as water vapor before it can fall to the ground as rain, one goal of this thesis is to understand if water vapor information can improve radar extrapolation techniques by giving the information needed to infer growth and decay. To do so, we use the GPS-Meteorology technique to measure the water vapor in the atmosphere and weather radar reflectivity to measure rainfall. By training a machine learning nowcasting algorithm using both variables and comparing its performance against a nowcasting algorithm trained on reflectivity alone, we draw conclusions as to the predictive power of adding water vapor information. Another goal of this thesis is to compare different machine learning techniques, viz., the random forest ensemble learning technique, which has shown success on a number of other weather prediction problems, and the current state-of-the-art machine learning technique for images and image sequences, convolutional neural network (CNN). We compare these in terms of problem representation, training complexity, and nowcasting performance. A final goal is to compare the nowcasting performance of our machine learning techniques against published results for current state-of-the-art model based nowcasting techniques.