Дисертації з теми "Satellite rainfall estimation"

Infrared (IR) imagery collected by geostationary satellites provides useful information about the dirunal evolution of cloud systems. These IR images can be analyzed to indicate the location of clouds as well as the pattern of cloud top temperatures (Tbs). During the past several decades, a number of different approaches for estimation of rainfall rate (RR) from Tb have been explored and concluded that the Tb-RR relationship is (1) highly nonlinear, and (2) seasonally and regionally dependent. Therefore, to properly model the relationship, the model must be able to: (1) detect and identify a non-linear mapping of the Tb-RR relationship; (2) Incorporate information about various cloud properties extracted from IR image; (3) Use feedback obtained from RR observations to adaptively adjust to seasonal and regional variations; and (4) Effectively and efficiently process large amounts of satellite image data in real -time. In this study, a kind of artificial neural network (ANN), called Modified Counter Propagation Network (MCPN), that incorporates these features, has been developed. The model was calibrated using the data around the Japanese Islands provided by the Global Precipitation Climatology Project (GPCP) First Algorithm Intercomparison Project (AIP-I). Validation results over the Japanese Islands and Florida peninsula show that by providing limited ground-truth observation, the MCPN model is effective in monthly and hourly rainfall estimation. Comparison of results from MCPN model and GOES Precipitation Index (GPI) approach is also provided in the study.

Infrared (IR) imagery collected by geostationary satellites provides useful information about the dirunal evolution of cloud systems. These JR images can be analyzed to indicate the location of clouds as well as the pattern of cloud top temperatures (Tbs). During the past several decades, a number of different approaches for estimation of rainfall rate (RR) from Tb have been explored and concluded that the Tb-RR relationship is (1) highly nonlinear, and (2) seasonally and regionally dependent. Therefore, to properly model the relationship, the model must be able to: (1) detect and identify a non-linear mapping of the Tb-RR relationship; (2) Incorporate information about various cloud properties extracted from IR image; (3) Use feedback obtained from RR observations to adaptively adjust to seasonal and regional variations; and (4) Effectively and efficiently process large amounts of satellite image data in real-time. In this study, a kind of artificial neural network (ANN), called Modified Counter Propagation Network (MCPN), that incorporates these features, has been developed. The model was calibrated using the data around the Japanese Islands provided by the Global Precipitation Climatology Project (GPCP) First Algorithm Intercompari son Project (AIP-I). Validation results over the Japanese Islands and Florida peninsula show that by providing limited ground-truth observation, the MCPN model is effective in monthly and hourly rainfall estimation. Comparison of results from MCPN model and GOES Precipitation Index (GPI) approach is also provided in the study.

Famennian (Upper Devonian) ammonoids and their biostratigraphy are reviewed with particular reference to the Sauerland and Oberfranken, West Germany. Host european species of the Order Clymeniida are described. The Famennian ammonoid zonal scheme is rationalised and within it 23 faunal levels are proposed. Ammonoid Zones and conodont zones are correlated, and the rhomboidea (conodont) Zone is newly recognised to be coeval with the curvispina Zone. The fOllowing genera and subgenera are dealt with in detail: Progonioclyrnenia, Endosiphonites, Sellaclymenia, Biloclymenia, Gonioclymenia (Gon.), Gonioclymenia (Kalloclymenia), Sphenocl~enia, Platycl~enia (Plat.), Plat. (Pleuroclymenia), Plat. (TrigonocIYmenia), Sulcoclymenia, Piricclymenia, Ornatoclymenia, Cyrto- clyymenia, Protactoclymenia, Carinoclymenia, Clymenia, protoxyclymenia , Kosmoclymenia, Genuclymenia, fymaclymenia, Genn. Nov. D, E, and F. In most cases the types of the type species are illustrated photographically for the first time. The following generic names are recognised to have been wrongly interpreted in the past, and, where necessary new names have been proposed. Kalloclymenia, Biloclymenia, Rectoclymenia and Falciclmenia. Two new subgenera and one new genus are proposed, and two generic names, Protactoclymenia and Endosiphonites, have been revived. Kosmoclymenia is split into four species groups by its ornament and Cymaclymenia bas been split into three species groups. Two widely used specific names are recognised to have been placed in the wrong genus; sedgwicki Munster is a pseudoclymenia (a goniatite), and ~pentina MUnster is a Protoxyclymenia.

The Ku band microwave remote sensor, SeaWinds, was developed at the National Aeronautics and Space Administration (NASA) Jet Propulsion Laboratory (JPL). Two identical SeaWinds instruments were launched into space. The first was flown onboard NASA QuikSCAT satellite which has been orbiting the Earth since June 1999, and the second instrument flew onboard the Japanese Advanced Earth Observing Satellite II (ADEOS-II) from December 2002 till October 2003 when an irrecoverable solar panel failure caused a premature end to the ADEOS-II satellite mission. SeaWinds operates at a frequency of 13.4 GHz, and was originally designed to measure the speed and direction of the ocean surface wind vector by relating the normalized radar backscatter measurements to the near surface wind vector through a geophysical model function (GMF). In addition to the backscatter measurement capability, SeaWinds simultaneously measures the polarized radiometric emission from the surface and atmosphere, utilizing a ground signal processing algorithm known as the QuikSCAT / SeaWinds Radiometer (QRad / SRad). This dissertation presents the development and validation of a mathematical inversion algorithm that combines the simultaneous active radar backscatter and the passive microwave brightness temperatures observed by the SeaWinds sensor to retrieve the oceanic rainfall. The retrieval algorithm is statistically based, and has been developed using collocated measurements from SeaWinds, the Tropical Rainfall Measuring Mission (TRMM) Microwave Imager (TMI) rain rates, and Numerical Weather Prediction (NWP) wind fields from the National Centers for Environmental Prediction (NCEP). The oceanic rain is retrieved on a spacecraft wind vector cell (WVC) measurement grid that has a spatial resolution of 25 km. To evaluate the accuracy of the retrievals, examples of the passive-only, as well as the combined active / passive rain estimates from SeaWinds are presented, and comparisons are made with the standard TRMM rain data products. Results demonstrate that SeaWinds rain measurements are in good agreement with the independent microwave rain observations obtained from TMI. Further, by applying a threshold on the retrieved rain rates, SeaWinds rain estimates can be utilized as a rain flag. In order to evaluate the performance of the SeaWinds flag, comparisons are made with the Impact based Multidimensional Histogram (IMUDH) rain flag developed by JPL. Results emphasize the powerful rain detection capabilities of the SeaWinds retrieval algorithm. Due to its broad swath coverage, SeaWinds affords additional independent sampling of the oceanic rainfall, which may contribute to the future NASA's Precipitation Measurement Mission (PMM) objectives of improving the global sampling of oceanic rain within 3 hour windows. Also, since SeaWinds is the only sensor onboard QuikSCAT, the SeaWinds rain estimates can be used to improve the flagging of rain-contaminated oceanic wind vector retrievals. The passive-only rainfall retrieval algorithm (QRad / SRad) has been implemented by JPL as part of the level 2B (L2B) science data product, and can be obtained from the Physical Oceanography Distributed Data Archive (PO.DAAC).
Ph.D.
School of Electrical Engineering and Computer Science
Engineering and Computer Science
Electrical Engineering PhD

Le plateau des Guyanes est une région qui est caractérisée à 90% d’une forêt tropicale primaire et compte pour environ 20% des réserves mondiales d’eau douce. Ce territoire naturel, au vaste réseau hydrographique, montre des intensités pluviométriques annuelles atteignant 4000 mm/an ; ce qui fait de ce plateau une des régions les plus arrosées du monde. De plus les précipitations tropicales sont caractérisées par une variabilité spatiale et temporelle importante. Outre les aspects liés au climat, l’impact des précipitations dans cette région du globe est important en termes d’alimentation énergétique (barrages hydroélectriques). Il est donc important de développer des outils permettant d’estimer quantitativement et qualitativement et à haute résolution spatiale et temporelle les précipitations dans cette zone. Cependant ce vaste espace géographique est caractérisé par un réseau de stations pluviométriques peu développé et hétérogène, ce qui a pour conséquence une méconnaissance de la répartition spatio-temporelle précise des précipitations et de leurs dynamiques.Les travaux réalisées dans cette thèse visent à améliorer la connaissance des précipitations sur le plateau des Guyanes grâce à l’utilisation des données de précipitations satellites (Satellite Precipitation Product : SPP) qui offrent dans cette zone une meilleure résolution spatiale et temporelle que les mesures in situ, au prix d’une qualité moindre en terme de précision.Cette thèse se divise en 3 parties. La première partie compare les performances de quatre produits d’estimations satellitaires sur la zone d’étude et tente de répondre à la question : quelle est la qualité de ces produits au Nord de l’Amazone et sur la Guyane française dans les dimensions spatiales et temporelles ? La seconde partie propose une nouvelle technique de correction de biais des SPP qui procède en trois étapes : i) utiliser les mesures in situ de précipitations pour décomposer la zone étudiée en aires hydro-climatiques ii) paramétrer une méthode de correction de biais appelée quantile mapping sur chacune de ces aires iii) appliquer la méthode de correction aux données satellitaires relatives à chaque aire hydro-climatique. On cherche alors à répondre à la question suivante : est-ce que le paramétrage de la méthode quantile mapping sur différentes aires hydro-climatiques permet de corriger les données satellitaires de précipitations sur la zone d’étude ? Après avoir montré l’intérêt de prendre en compte les différents régimes pluviométriques pour mettre en œuvre la méthode de correction QM sur des données SPP, la troisième partie analyse l’impact de la résolution temporelle des données de précipitations utilisées sur la qualité de la correction et sur l’étendue spatiale des données SPP potentiellement corrigeables (données SPP sur lesquelles la méthode de correction peut s’appliquer avec efficacité). Concrètement l’objectif de cette partie est d’évaluer la capacité de notre méthode à corriger sur une large échelle spatiale le biais des données TRMM-TMPA 3B42V7 en vue de rendre pertinente l’exploitation de ce produit pour différentes applications hydrologiques.Ce travail a permis de corriger les séries satellites journalières à haute résolution spatiale et temporelle sur le plateau des Guyanes selon une approche nouvelle qui utilise la définition de zones hydro-climatiques. Les résultats positifs en terme de réduction du biais et du RMSE obtenus grâce à cette nouvelle approche, rendent possible la généralisation de cette nouvelle méthode dans des zones peu équipées en pluviomètres
The Guiana Shield is a region that is characterized by 90% of a primary rainforest and about 20% of the world’s freshwater reserves. This natural territory, with its vast hydrographic network, shows annual rainfall intensities up to 4000 mm/year; making this plateau one of the most watered regions in the world. In addition, tropical rainfall is characterized by significant spatial and temporal variability. In addition to climate-related aspects, the impact of rainfall in this region of the world is significant in terms of energy supply (hydroelectric dams). It is therefore important to develop tools to estimate quantitatively and qualitatively and at high spatial and temporal resolution the precipitation in this area. However, this vast geographical area is characterized by a network of poorly developed and heterogeneous rain gauges, which results in a lack of knowledge of the precise spatio-temporal distribution of precipitation and their dynamics.The work carried out in this thesis aims to improve the knowledge of precipitation on the Guiana Shield by using Satellite Precipitation Product (SPP) data that offer better spatial and temporal resolution in this area than the in situ measurements, at the cost of poor quality in terms of precision.This thesis is divided into 3 parts. The first part compares the performance of four products of satellite estimates on the study area and attempts to answer the question : what is the quality of these products in the Northern Amazon and French Guiana in spatial and time dimensions ? The second part proposes a new SPP bias correction technique that proceeds in three steps: i) using rain gauges measurements to decompose the studied area into hydro climatic areas ii) parameterizing a bias correction method called quantile mapping on each of these areas iii) apply the correction method to the satellite data for each hydro-climatic area. We then try to answer the following question : does the parameterization of the quantile mapping method on different hydro-climatic areas make it possible to correct the precipitation satellite data on the study area ? After showing the interest of taking into account the different rainfall regimes to implement the QM correction method on SPP data, the third part analyzes the impact of the temporal resolution of the precipitation data used on the quality of the correction and the spatial extent of potentially correctable SPP data (SPP data on which the correction method can be applied effectively). In summary, the objective of this section is to evaluate the ability of our method to correct on a large spatial scale the bias of the TRMM-TMPA 3B42V7 data in order to make the exploitation of this product relevant for different hydrological applications.This work made it possible to correct the daily satellite series with high spatial and temporal resolution on the Guiana Shield using a new approach that uses the definition of hydro-climatic areas. The positive results in terms of reduction of the bias and the RMSE obtained, thanks to this new approach, makes possible the generalization of this new method in sparselygauged areas

Depuis le développement des mesures satellites de nombreuses missions spatiales sont dédiées au suivi de l'atmosphère et de la surface terrestre. Ces travaux de thèse s'inscrivent dans le cadre de la mission Megha-Tropiques dédiée au cycle de l'eau et de l'énergie en zone tropicale. L'objectif est d'évaluer le potentiel des estimations de précipitation par satellite pour des applications hydrologiques en zone tropicale. Les Tropiques réunissent les plus grands fleuves du globe, mais ne bénéficient pas de réseaux d'observation in-situ denses et continus permettant une gestion intégrée efficace de la ressource et des systèmes d'alertes. Les estimations des précipitations issues des systèmes d'observation satellite offrent une alternative pour ces bassins peu ou pas instrumentés et souvent exposés aux extrêmes climatiques. C'est le cas du fleuve Niger, qui a subi une grande variabilité climatique depuis les années 1950, mais aussi d'importants changements environnementaux et hydrologiques. Depuis les années 2000, le Niger moyen connaît une recrudescence des inondations pendant la période de crue Rouge (engendrée par ses affluents sahéliens pendant la mousson). A Niamey, des niveaux record de hauteur d'eau et de période d'inondation ont été enregistrés en 2003, 2010, 2012 et 2013, engendrant de nombreuses pertes humaines et matérielles. Ces travaux analysent l'influence du forçage pluviométrique sur les inondations liées à la crue Rouge à Niamey. Une gamme de produits pluviométriques (in situ et satellite) et la modélisation hydrologique (ISBA-TRIP) sont combinés pour étudier : (i) l'apport des produits satellite pour diagnostiquer la crue Rouge récente, (ii) l'impact des caractéristiques des produits et de leurs incertitudes sur les simulations et enfin (iii) l'évaluation du rôle des précipitations, face aux changements de conditions de surface, dans l'évolution de la crue Rouge à Niamey depuis les années 1950. L'étude a mis en évidence l'impact des caractéristiques des estimations des précipitations (cumul, intensité et distribution spatio-temporelle) sur la modélisation hydrologique et le potentiel des produits satellites pour le suivi des inondations. Les caractéristiques des précipitations se propageant dans la modélisation, la détection des inondations est plus efficace avec une approche relative à chaque produit plutôt qu'avec un seuil absolu. Ainsi des produits présentant des biais peuvent être envisagés pour la simulation hydrologique et la détection des inondations. Le nouveau produit TAPEER de la mission MT présente un fort potentiel hydrologique, en 2012 et pour la zone d'étude. D'autre part, l'étude de la propagation de l'erreur associée à ces précipitations a mis en évidence, la nécessité de déterminer la structure du champ d'erreur pour l'utilisation d'une telle information en hydrologie. Enfin la modélisation a été utilisée comme levier pour décomposer les sensibilités de la crue Rouge aux variations des précipitations et des conditions de surface. Pour simuler les changements hydrologiques entre les périodes 1953-1982 et 1983-2012, les changements d'occupation du sol et d'aire de drainage doivent être pris en compte. Puis les variations des précipitations peuvent expliquer les changements majeurs décennaux et annuels entre les années 1983 et 2012
Since the development of satellite based remote sensing in the 1970s, many missions have been dedicated to monitoring the terrestrial atmosphere and surfaces. Some of these satellites are dedicated to the Tropics with specific orbits. Megha-Tropiques (MT) is devoted to the water and energy cycle in the tropical atmosphere and provides an enhanced sampling for rainfall estimation in the tropical region. This PhD work was initiated within MT hydro-meteorological activities, with the objective of assessing the hydrological potential of satellite rainfall products in the Tropics. The world most important rivers lay in tropical areas where the in situ observation networks are deficient. Alternative information is therefore needed for water resource management and alert systems. The present work focuses on the Niger River a basin which has undergone drastic climatic variations leading to disasters such as droughts and floods. Since 1950, the Niger has been through 3 main climatic periods: a wet period (1950-1960), a long and intense drought period (1970-1980) and since 1990 a partial recovery of the rainfall. These climatic variations and the anthropic pressure, have modified the hydrological behaviour of the basin. Since 2000, the middle Niger River has been hit by an increase of floods hazards during the so-called Red flood period. In Niamey city, the highest river levels and the longest flooded period were recorded in 2003, 2010, 2012 and 2013, leading to heavy casualties and property damage. This study combines hydrological modelling and a variety of rainfall estimation products (satellite and in-situ) to meet several objectives: (i) the simulation of the Niamey Red flood and the detection of floods (during the recent period 2000-2013) (ii) the study of the propagation of satellite rainfall errors in hydrological modelling (iii) the evaluation of the role of rainfall variability, and surface conditions, in the changes of the Red flood in Niamey since the 50s. The global model ISBA-TRIP, is run with a resolution of 0.5° and 3h, and several rainfall products were used as forcing. Products derived from gauges (KRIG, CPC), pure satellite products (TAPEER, 3B42RT, CMORPH, PERSIANN) and mixed satellite products adjusted by rain gauges (3B42v7, RFE2, PERSIANN-CDR). This work confirms the hydrological potential of satellite rainfall products and proposes an original approach to overcome their biases. It highlights the need for documenting the errors associated with the rainfall products and the error structure. Finally, the hydrological modelling results since the 1950s have given a new understanding of the relative role of rainfall and surface conditions in the drastic increase of flood risk in Niamey

Most infrared-based techniques of satellite rainfall estimation contain substantial uncertainties due to the indirect relationship between precipitation particles and space-borne infrared observations of clouds. Generally, these uncertainties include (1) IR temperature threshold defining cold clouds; (2) inclusion of no-rain clouds; (3) exclusion of warm rain clouds; and (4) the coefficients between rain rate and cloud-top properties. To address these uncertainties, a methodology, Cloud Patch Analysis, was developed to estimate rainfall by removing large portion of no-rain clouds from IR cloud imagery. Seven cloud features, including physical, geometric and textural, were defined, and ID3, an inductive decision tree, was used to identify no-rain clouds. Particularly, textural characteristics were extended from square images to irregular cloud patches to extract cloud features related to rainfall. In addition, the method adopted a mechanism to adjust IR temperature threshold according to locations and seasons, and this adjustment can be made by the combination of microwave observations by polar-orbiting satellites with infrared observations by geostationary satellites. The application of the adjusted IR threshold to GPI algorithm showed significant improvement for monthly rainfall estimation. The method was applied to the Japanese Islands and surrounding oceanic regions in June and July/August 1989 and to the Florida region in June and August 1996. The monthly rainfall estimates by the proposed method showed significant and consistent improvements over those by GPI.

Satellite information, in combination with conventional point source measurements, can be a valuable source of information. This thesis is devoted to the spatial estimation of areal rainfall over a region using both the measurements from a dense and sparse network of rain-gauges and images from the meteorological satellites. A primary concern is to study the effects of such satellite assisted rainfall estimates on the performance of rainfall-runoff models. Low-cost image processing systems and peripherals are used to process and manipulate the data. Both secondary as well as primary satellite images were used for analysis. The secondary data was obtained from the in-house satellite receiver and the primary data was obtained from an outside source. Ground truth data was obtained from the local Water Authority. A number of algorithms are presented that combine the satellite and conventional data sources to produce areal rainfall estimates and the results are compared with some of the more traditional methodologies. The results indicate that the satellite cloud information is valuable in the assessment of the spatial distribution of areal rainfall, for both half-hourly as well as daily estimates of rainfall. It is also demonstrated how the performance of the simple multiple regression rainfall-runoff model is improved when satellite cloud information is used as a separate input in addition to rainfall estimates from conventional means. The use of low-cost equipment, from image processing systems to satellite imagery, makes it possible for developing countries to introduce such systems in areas where the benefits are greatest.

Esta tesis explora la precisión de dos métodos de estimación de precipitación, Auto-Estimator y CRR (Convective Rainfall Rate), generados a partir de imágenes infrarrojas y visibles del Meteosat. Ambos métodos junto con una serie de correcciones de la intensidad de lluvia estimada se aplican y se verifican en dos casos de inundaciones acaecidas en zonas mediterráneas. El primer caso ocurrió en Albania del 21 al 23 de septiembre de 2002 y el segundo, conocido como caso Montserrat, ocurrió en Cataluña la noche del 9 al 10 se junio de 2000. Por otro lado se investiga la posibilidad de realizar calibraciones de ambos métodos directamente con datos de estaciones pluviométricas cuando lo común es calibrar con datos de radares meteorológicos. También se propone cambios en algunas de las correcciones ya que parecen mejorar los resultados y se propone una nueva corrección muy eficiente que utiliza las descargas eléctricas para determinar la zonas más convectivas y de mayor precipitación de los sistemas nubosos.
This Thesis work explores the precision of two methods to estimate rainfall called Auto-Estimator and CRR (Convective Rainfall Rate). They are obtained by using infrared and visible images from Meteosat. Both Algorithms within a set of correction factors are applied and verified in two severe flood cases that took place in Mediterranean regions. The first case has occurred in Albania from 21 to 23 September 2002 and the second, known as the Montserrat case, has occurred in Catalonia the night from the 9 to 10 of June 2000. On the other hand it is explored new methods to perform calibrations to both satellite algorithms using direct rain rates from rain gauges. These kinds of adjustments are usually done using rain rates from meteorological radars. In addition it is proposed changes on some correction factors that seem to improve the results on estimations and it is defined an efficient correction factor that employ electrical discharges to detect the most convective and rainy areas in cloud systems.

As estimativas de chuva por meio do sensoriamento remoto são, atualmente, fonte potencialmente útil para as mais diversas aplicações hidrológicas e climatológicas, especialmente em regiões onde as medições convencionais são escassas, como a Amazônia. Neste trabalho, foram analisadas as estimativas de chuva por satélite como variáveis de entrada ao modelo hidrológico MGB-IPH (Collischonn, 2001). Este modelo simula o ciclo hidrológico através das relações físicas e conceituais de todo processo, sendo os produtos de chuva por satélite avaliados o 3B42, 3B42RT e CMORPH. A primeira área de estudo é a bacia do rio Huallaga localizada dentro do território do Peru, região caracterizada por ter uma topografia complexa e pertencente a uma das nascentes do rio Amazonas. A segunda avaliação foi feita para a bacia do rio Amazonas, sendo esta caracterizada por ter uma grande variabilidade climatológica a diferentes altitudes e regimes hidrológicos diferentes, além de uma pobre distribuição de postos pluviométricos. No caso da bacia do rio Huallaga foram realizadas comparações da chuva média estimada por satélite com observada em intervalos de tempo diário, mensal, sazonal e anual. Estes resultados mostram que os produtos 3B42 e CMORPH subestimam valores médios da bacia comparada com chuva média ponderada por pluviômetros. Na simulação da bacia do rio Huallaga se efetuaram calibrações dos parâmetros para cada fonte de chuva resultando com melhor ajuste de vazões máximas para o produto CMORPH e pior ajuste para o produto 3B42, estes ajustes melhoraram para a chuva do produto CMORPH corrigido com estações pluviométricas. Por outra parte, no caso de análises da bacia do rio Amazonas, foi calculada a chuva média anual para os três produtos de satélite (3B42, 3B42RT e CMORPH), os resultados mostraram maior chuva média a favor de CMORPH, seguido de 3B42RT e finalmente o produto 3B42. A simulação da bacia do rio Amazonas mostrou melhores coeficientes de Nash-Sutcliffe com o produto 3B42 em várias estações do Brasil. Com o produto 3B42RT mostram melhores coeficientes nas estações localizadas na rede principal do rio Amazonas, e com o produto CMORPH mostrou melhores coeficientes em algumas estações como na bacia dos rios Tapajós (Brasil) e Urubamba (Peru).
Currently, satellite rainfall estimates using remote sensing are a potential source of information for hydrological and climatological applications. It applies mainly for regions where conventional measurements are scarce such as the Amazon Basin. In this work, the satellite rainfall estimates were analyzed as input variables to the hydrological model MGBIPH (Collischonn, 2001). This model simulates the hydrological cycle through physical and conceptual relationships where products 3B42, 3B42RT and CMORPH are evaluated. The first evaluation case corresponds to the Huallaga basin located in Peru, being one of the current Amazon highlands characterized by a complex topography. The second evaluation case corresponds to the Amazon basin characterized by a great climatological variability at different altitudes, different hydrological regimes and poor distributions of raingauges. In the case of the Huallaga River basin, comparisons were made between the estimated average satellite rainfall and the observed rainfall for different intervals of time (daily, monthly, seasonal and annual). These results show that the products 3B42 and CMORPH underestimate the basin average rainfall when compared with the weighted average of raingauge measurements. During the Huallaga basin simulation, calibrations of some parameters for each rainfall data were realized. Obtaining the best and worst fitting results with the CMORPH and 3B42 products for the case of maximum discharges, respectively. This rainfall fitting improves for the CMORPH product when raingauge corrections are included. On the other hand, the annual average rainfall value was obtained for each satellite product (3B42, 3B42RT e CMORPH) for the analysis of the Amazon basin. In this calculation, the greater results for the annual average rainfall values are obtained in the following order CMORPH, 3B42RT and 3B42. Moreover, this simulation seems to yield best Nash-Sutcliffe coefficients for the 3B42 product for various Brazilian stations. For stations located in the main stream of the Amazon River the Nash-Sutcliffe coefficients obtained with the 3B42RT product are the best. The CMORPH product yield the best coefficients for the stations located in Tapajós (Brazil) and Urubamba (Peru) basin.

博士
國立臺灣大學
生物環境系統工程學系暨研究所
91
In recent years, rainfall estimation using weather satellite imagery has gained increasing attention. GMS imagery and hourly rainfall data of eleven typhoons during 1997-2000 are used in this study. The correlation between cloud top temperature (CCT) derived from weather satellite imagery and corresponding ground measurements of hourly rainfall are analyzed at different spatial and temporal scales. Two approaches are used to estimate rainfall in this study: (i) image classification method and (ii) spatial and temporal rainfall apportionment scheme (STRAS). After the models are established, the cross validation scheme is used to validate the stability and feasibility of the model. Finally, real-time forecasting of 5-km and watershed scale is accomplished by the kalman filtering algorithm. Preliminary results show the relationship between CTT and rain rate is dependent on individual typhoon event magnitude and total rainfall depth. The correlation between CCT and rain rate at the same time becomes better as the spatial scale is larger. The correlation of the average CCT and next three to six hour accumulative rainfall in basin scale also becomes better as the time frame is longer. In image classification approach, the maximum likelihood and bayesian classification method are used to analyze the data. The total accuracy of two methods using 14 factor scores as classification features are 87.068% and 88.321%, respectively. In STRAS approach, the correlation between raingauge point measurements and rainfall estimates by spatial convolution using IR1 ,IR1 and IR3 , and all three infrared images are 0.742, 0.858, and 0.932, respectively. Correlation coefficients increase to 0.874, 0.940, and 0.970 respectively, when pixel-average rainfall estimates are considered by block kriging. The correlations also become better when the spatial and temporal scale increase. Results of a cross validation scheme reveal that the correlation coefficients are over 0.7 in most raingauges. The result suggests a great potential in real-time rainfall forecasting using satellite images. The testing result of the rainfall forecasting by the kalman filtering algorithm demonstrates that the proposed model can master the trend of rainfall variation. However, the statistical characteristics of some parameters in kalman filtering algorithm are yet to be verified.

碩士
淡江大學
水資源及環境工程學系碩士班
95
The main purpose of this study is to explore the influence of satellite imagery information on rainfall estimation using artificial neural networks. However, it is often difficult to extract interpretable information from satellite images, as data dimensions are large and nonlinear. We proposed the self-organizing map (SOM), one of artificial neural network adept at pattern cognition. In this study, watershed rainfall estimation models are constructed to forecast the rainfall summation of future six hours during typhoon events. The models are based on SOM or linear regression to investigate the characteristics of satellite imagery information and its influence on rainfall estimation. The available data are hourly rainfall data of sixteen rainfall gauge stations in the Shihmen watershed from 25 typhoon events and GMS-5/MTSAT remotely sensed data are collected from 2000 to 2004 and 2006. In order to investigate the characteristics and compare the performance among the different models, we designed three cases with different sizes or amount of rainfall in training data, then constructed six different models, multivariate linear regression model (MLR), back-propagation neural network (BP), self-organizing map linking with BP (SOMBP), self-organizing map linking with linear regression (SOMMLR), SOMBP linking with BP (SOMBP+BP) and SOMMLR linking with BP linear regression (SOMMLR+BP), to estimate the future six-hour rainfall summation. The input variables have two types: the past three six-hour rainfall summations and satellite images. The results show that (1) the MLR models have nice performances when the input variable only include the past rainfall summations, (2) SOM indeed has the ability to extract patterns from satellite data, (3) SOMBP and SOMMLR can get better results when the input variables are the past rainfall summations and satellite images. The satellite imagery information is indeed helpful to improve the accurate of rainfall estimation.

碩士
淡江大學
水資源及環境工程學系碩士班
98
The main purpose of this study is to explore the influence of satellite imagery and meteorological data on typhoon rainfall forecast using artificial neural networks. The self-organizing map (SOM) is adept at recognizing infrared and visible images and can extract some useful information. In this study, six watershed rainfall estimation models are constructed to forecast the amount of rainfall for one, three and six-hour totals during typhoon events. The models are based on SOM, back-propagation neural network (BPNN) or linear regression to investigate the characteristics of satellite imagery information and its influence on rainfall forecast. Twenty-seven typhoon events are collected from 2000 to 2007. The available data are GMS-5/MTSAT remotely sensed data, hourly rainfall data of sixteen rainfall gauge stations of the Shihmen watershed, wind velocity and atmospheric pressure data of three meteorological observation stations. In order to investigate the characteristics and compare the performance among the different models, we design different cases for forecasting the rainfall totals in the daytime and the whole day. Six different models, multivariate linear regression model (MLR), back-propagation neural network (BP), self-organizing map linking with BP (SOMBP), self-organizing map linking with linear regression (SOMMLR), SOMBP linking with BP (SOMBPI+BP) and SOMMLR linking with BP linear regression (SOMMLRI+BP), are constructed to forecast rainfall totals. Seven different combinations of the inputs are used to investigate the effect of rainfall forecast. The results show that (1) the MLR and BP models have nice performances when the input variable only include the past rainfall totals of gauge stations, (2) SOM indeed has the ability to extract patterns from satellite data, (3) SOM can improve results when the rainfall totals are joined, (4) the wind velocity and atmospheric pressure data are helpless for rainfall forecast. The satellite imagery information is indeed helpful to improve the accurate of rainfall forecast.

The European Union's Earth Observation program, Copernicus, aims to create a vast amounts of global data from satellites and ground-based, airborne and seaborne measurement systems with the goal of providing information to help service providers, public authorities, and other international organizations improve European citizens' quality of life. With the aim of reaching this goal, a new family of missions called Sentinels has been developed by the European Spatial Agency, ESA, specifically for the operational needs of the program. These missions carry a range of cutting-edge technologies, such as radars and multi-spectral imaging instruments, for land, ocean and atmospheric monitoring. Multiple kinds of high-resolution data are now available to the scientific community, which is working to adapt and develop the existing models and algorithms to the new information. With this objective, two of the most important variables that contribute to the water cycle, rainfall and river discharge, were selected in this thesis to be estimated by the use of the information obtained from Sentinel-1 and Sentinel-2 sensors. The monitoring of these two variables is fundamental in many hydrological applications, like flood and landslide forecasting and water resources management, and their impact is clearly visible from space. In-situ measurements are the traditional data source of them, but the worldwide declining number of stations, their low spatial density and the data access problem limit their use. Satellite sensors have been therefore adopted to support and, in some cases, substitute the existing gauge network in estimating river discharge and rainfall, thanks to the strong growth in technologies and applications. Two valuable examples of this are SM2RAIN algorithm, which allows to estimate rainfall from Soil Moisture (SM) observations by exploiting the inversion of the soil water balance equation, and the CM approach, a non-linear regression model capable of linking the ground measurements of river discharge to the near-infrared (NIR) reflectance ratio between a dry and a wet pixel chosen around the border of a river. Notwithstanding their usefulness, until now several limitations affected these two methodologies. The main issue with satellite derived rainfall data was their low spatial resolution which could not overcome 10 km, a quantity insufficient to obtain accurate information over many areas and posing important constraint on their use for many applications and fields, which require more and more detailed information. Similarly, the resolution of the available NIR data was not suitable to provide information for narrow rivers (< 250 m wide), nor to study river features and patterns that were here averaged within a single pixel. The recently available high-resolution data from the Sentinel Missions of the Copernicus program offer an opportunity to overcome these issues. The data from Sentinel-1 mission can be used to obtain a high spatial resolution SM product, named S1-RT1, that is adopted in this thesis to derive 1 km spatial resolution (500 m spacing) rainfall data over the Po River basin from it, through the algorithm SM2RAIN. The rainfall derived from the 25 km ASCAT SM product (12.5 km spacing), resampled to the same grid of S1-RT1, is compared to the latter to evaluate the potential benefits of such product. SM2RAIN algorithm needs to be calibrated against a benchmark, which poses important limitations on the applicability of the analysis in data scarce regions. In order to overcome this issue, a parameterized version of SM2RAIN algorithm is previously developed relying on globally distributed data, to be used along with the standard approach in the high-resolution rainfall estimation. The performances of each obtained product are then compared, to assess both the parameterized SM2RAIN capabilities in estimating rainfall and the benefits deriving from Sentinel-1 high spatial resolution. For the river discharge estimation, the use of Sentinel-2 NIR reflectances within the CM approach is investigated to support the hypothesis that a higher satellite product’s spatial resolution, i.e., 10 m (vs. a medium-resolution, i.e., 250 m), is able to better identify the periodically flooded pixels, more related to the river dynamics, with obvious advantages for river discharge estimation. Moreover, the improved resolution allows both a finer distinction between vegetation, soil and water and the characterization of water turbidity in the river area, which is important to correctly estimate the river discharge using this approach. A new formulation enriched by the sediment component is proposed along with a procedure to localize the periodically flooded pixels without the intake of calibration data, which is a first step towards a completely uncalibrated procedure for the river discharge estimation, fundamental for ungauged rivers. The obtained results show that the high-resolution information from Copernicus actually increase the accuracy of the satellite derived products. Good estimates of rainfall are obtainable from Sentinel-1 when considering aggregation time steps greater than 1 day, since to the low temporal resolution of this sensor (from 1.5 to 4 days over Europe) prevents its application to infer daily rainfall. In particular, the rainfall estimates obtained from Sentinel-1 sensors outperform those from ASCAT in specific areas, like in valleys inside mountain regions and most of the plains, confirming the added value of the high spatial resolution information in obtaining spatially detailed rainfall. The use of a parameterized version of SM2RAIN produces performances similar to those obtained with SM2RAIN calibration, attesting the reliability of the parameterized algorithm for rainfall estimation in this area and fostering the possibility to apply SM2RAIN worldwide even without the availability of a rainfall benchmark product. Similarly, the river discharge estimation from Sentinel-2 reflectances from selected stations along two Italian rivers, the Po and the Tiber, confirms that reliable performance can be obtained from high-resolution imagery. Specifically, over both the stations the new formulation improves the river discharge accuracy and over the Po River the best performances are obtained by the uncalibrated procedure. Google Earth Engine (GEE) platform has been employed for the data analysis, allowing to avoid the download of big amounts of data, fostering the reproducibility of the analysis in different locations. Il programma per l’Osservazione della Terra dell’Unione Europea, Copernicus, mira a creare una grande quantità di dati globali da satelliti e sistemi di misurazione terrestri, aerei e marittimi con l'obiettivo di fornire informazioni per assistere i fornitori di servizi, le autorità pubbliche e altre organizzazioni internazionali a migliorare la qualità della vita dei cittadini Europei. Per raggiungere questo obiettivo, l’Agenzia Spaziale Europea (ESA) ha sviluppato una nuova famiglia di missioni satellitari, denominate Sentinel, progettata specificatamente per le esigenze operative del programma. Queste missioni trasportano una gamma di tecnologie all'avanguardia per il monitoraggio terrestre, oceanico e atmosferico, come radar e strumenti di scansione multispettrale. Diversi tipi di dati ad alta risoluzione sono ora disponibili per la comunità scientifica, che sta lavorando per adattare e sviluppare i modelli e gli algoritmi esistenti alle nuove informazioni. Per questa ragione, due delle variabili più importanti del ciclo dell'acqua, la precipitazione e la portata fluviale, sono state selezionate in questa tesi per essere stimate attraverso le informazioni ottenute dai sensori Sentinel-1 e Sentinel-2. Queste variabili sono state scelte perché il loro monitoraggio è fondamentale in molte applicazioni idrologiche, come la previsione di alluvioni e frane e la gestione delle risorse idriche, e inoltre il loro impatto è chiaramente visibile dallo spazio. La fonte tradizionale di questi dati sono le stazioni di misura in situ, ma il decrescente numero dei sensori operativi, la loro ridotta rappresentatività spaziale e i problemi relativi all’accesso e alla condivisione dei dati ne ostacolano l’uso a livello globale. I sensori satellitari sono stati quindi adottati come supporto o, in alcuni casi, alternativa alla rete di misura esistente per la stima delle precipitazioni e delle portate fluviali, grazie alla costante evoluzione delle tecnologie impiegate e della ricerca sulle loro applicazioni. Due validi esempi in tal senso sono l'algoritmo SM2RAIN e l'approccio CM: il primo permette di stimare le precipitazioni in un’area a partire da una serie temporale di umidità del suolo della stessa (ottenibile da satellite), grazie all'inversione dell'equazione di bilancio idrico del suolo; il secondo è un modello di regressione non lineare capace di ottenere una stima della portata fluviale attraverso l’uso di sensori satellitari sensibili alla radiazione del vicino infrarosso (NIR), grazie alle differenze di riflettanza tra l’acqua e il terreno in questa regione dello spettro elettromagnetico. Nonostante la loro utilità, fino ad ora queste due metodologie hanno mostrato diversi limiti. Il problema principale con i dati di precipitazione derivati da satellite è la loro bassa risoluzione spaziale che non eccede i 10 km, una quantità insufficiente per ottenere informazioni accurate in molte regioni del mondo e che pone un importante vincolo al loro utilizzo in diversi campi e tipi di applicazione, richiedenti invece un grado di dettaglio sempre maggiore. Allo stesso modo, la risoluzione dei dati satellitari NIR disponibili non è sufficiente a fornire informazioni per fiumi stretti (< 250 m di larghezza), né a studiare le caratteristiche e i dettagli dei fiumi che vengono invece mediati all'interno di un singolo pixel. La recente disponibilità di dati ad alta risoluzione delle missioni Sentinel del programma Copernicus, però, offre l'opportunità di superare questi problemi. Le informazioni ottenute dalla missione Sentinel-1 possono essere infatti usate per ottenere un prodotto di umidità del suolo ad alta risoluzione spaziale, chiamato S1-RT1, che è stato selezionato in questa tesi per derivarne dati di pioggia a 1 km di risoluzione spaziale (500 m di spaziatura) per il bacino del fiume Po, attraverso l’utilizzo dell’algoritmo SM2RAIN. I benefici potenziali di tale prodotto sono stati valutati attraverso il confronto della pioggia ottenuta con quella derivata dal prodotto satellitare di umidità del suolo ASCAT, caratterizzato da 25 km di risoluzione spaziale (12.5 km di spaziatura) e opportunatamente ricampionato sulla griglia di S1-RT1. L’algoritmo SM2RAIN ha bisogno di essere calibrato attraverso l’uso di un prodotto di riferimento, cosa che pone importanti limiti al suo utilizzo in regioni con scarsa disponibilità di dati osservati. Per superare questo problema, è stata quindi sviluppata una versione parametrizzata dell’algoritmo SM2RAIN indipendente dai dati osservati, da affiancare alla versione standard per la stima della precipitazione ad alta risoluzione. Le performance di ciascun prodotto ottenuto sono state quindi confrontate, in modo da valutare sia le capacità del prodotto parametrizzato di stimare la pioggia senza il supporto di dati di calibrazione, sia i benefici derivanti dall’uso dei dati ad alta risoluzione spaziale di Sentinel-1. Per quanto riguarda la stima della portata fluviale, invece, l’approccio CM è stato applicato alle immagini NIR ottenute dal sensore Sentinel-2 per verificare l’ipotesi che una migliore risoluzione spaziale del prodotto satellitare adottato (10 m di Sentinel-2 contro una risoluzione media di 250 m dei suoi predecessori) sia capace di meglio identificare i pixel periodicamente allagati e quindi sensibili alle variazioni della portata fluviale, con conseguenti benefici alla stima di quest’ultima. La maggiore risoluzione rende inoltre possibile una più accurata distinzione tra vegetazione, suolo e acqua, nonché la caratterizzazione della torbidità dell’acqua nel tratto di fiume selezionato, fattori importanti per stimare correttamente la portata fluviale usando quest’approccio. È stato dunque possibile introdurre una nuova formulazione dell’approccio CM arricchita della componente dei sedimenti, nonché una procedura per la localizzazione dei pixel periodicamente allagati senza l’utilizzo di dati di calibrazione, che rappresenta un primo passo verso una procedura completamente non calibrata per la stima della portata fluviale, fondamentale in fiumi non strumentati. I risultati ottenuti mostrano che le informazioni ad alta risoluzione provenienti dal programma Copernicus possono effettivamente migliorare l’accuratezza dei prodotti di pioggia e portata fluviale ottenuti da satellite. È possibile ottenere stime attendibili di pioggia da Sentinel-1 quando tempi di aggregazione maggiori di un giorno sono presi in considerazione, dato che la ridotta risoluzione temporale del sensore (da 1.5 a 4 giorni per l’Europa) ne previene l’applicazione per l’ottenimento della pioggia a risoluzione giornaliera. In particolare, le stime della pioggia ottenute dai sensori Sentinel-1 hanno prestazioni migliori di quelle di ASCAT in aree specifiche, come le valli all’interno delle regioni montuose e la maggior parte delle pianure, confermando il valore aggiunto dalla elevata risoluzione spaziale nell’ottenimento di un prodotto di pioggia spazialmente dettagliato. L’uso della versione parametrizzata di SM2RAIN mostra prestazioni molto simili a quelle ottenute dalla calibrazione dello stesso, dimostrando l’affidabilità dell’algoritmo parametrizzato per la stima della pioggia nell’area considerata e la possibilità di applicare SM2RAIN in tutto il mondo anche senza la disponibilità di un prodotto di pioggia osservata di riferimento. Similmente, la stima della portata fluviale a partire dalla riflettanza misurata da Sentinel-2 per le stazioni selezionate lungo due fiumi italiani, il Po e il Tevere, conferma che buone prestazioni possono essere ottenute dall’utilizzo di immagini satellitari ad alta risoluzione. Specificatamente, la nuova formulazione permette di migliorare l’accuratezza della portata fluviale stimata in entrambe le stazioni, e, per il fiume Po, le migliori prestazioni sono ottenute dall’uso della procedura non calibrata, provandone la validità. Va infine sottolineato l’impiego della piattaforma Google Earth Engine per l’analisi dei dati di Sentinel-2, che ha permesso di evitare lo scaricamento di ingenti quantità di dati, favorendo la riproducibilità delle analisi anche in diverse località.

碩士
國立中央大學
大氣科學學系
106
The Tropical Rainfall Potential (TRaP) technique presented by Kidder et al. in 2005, shifting rainfall distribution from satellite retrieval, and forecasting rainfall for tropical cyclone. Chen(2010) improved TRaP rainfall forecast practicality by adding orographic effect with historical rainfall distribution(I-TRaP). Since I-TRaP forecast uses rainfall distribution from satellite, how to get better rainfall distribution is an important issue. There is only single satellite rainfall product in past study, limited by temporal resolution. For many study, The performance of multi-satellite rainfall products with high spatial-temporal resolution(0.1°-0.25°, 0.5-3h) are getting better recently but less discussed on heavy rainfall especially for typhoon. This study compares few common multi-satellite products (GSMaP, IMERG, PERSIANN) with typhoon heavy rainfall in the North-West Pacific, GSMaP is better. There are different performance between convective and stratiform rainfall. Indeed, the PMW retrieval fail to classification in rainfall type determination during microwave rainfall retrieving, but not cause rainfall error. In addition, compare liquid water content and rainfall error, the PMW retrieval still cannot estimate liquid water accurately in moderate to heavy rainfall. Apply GSMaP to I-TRaP and calculate typhoon rainfall forecast over Taiwan. In order to highlight satellite rainfall distribution, modify earlier method only revising total rainfall and using historical rainfall distribution, calculate rainfall regression by individual point. This method will predict more heavy rainfall but more false alarm. Compare earlier I-TRaP using SSMIS, GSMaP with high spatial-temporal resolution is more useful for I-TRaP forecast, and more prediction of heavy rainfall.

Error characteristics associated with satellite-derived precipitation products are important for atmospheric and hydrological model data assimilation, forecasting, and climate diagnostic applications. This information also aids in the refinement of physical assumptions within algorithms by identifying geographical regions and seasons where existing algorithm physics may be incorrect or incomplete. Examination of relative errors between independent estimates derived from satellite microwave data is particularly important over regions with limited surface-based equipments for measuring rain rate such as the global oceans and tropical continents. In this context, analysis of microwave based satellite datasets from the Tropical Rainfall Measuring Mission (TRMM) enables to not only provide information regarding the inherent uncertainty within the current TRMM products, but also serves as an opportunity to prototype error characterization methodologies for the TRMM follow-on program, the Global Precipitation Measurement (GPM) . Most of the TRMM uncertainty evaluation studies focus on the accuracy of rainfall accumulated over time (e.g., season/year). Evaluation of instantaneous rainfall intensities from TRMM orbital data products is relatively rare. These instantaneous products are known to potentially cause large uncertainties during real time flood forecasting studies at the watershed scale. This is more so over land regions, where the highly varying land surface emissivity offers a myriad of complications, hindering accurate rainfall estimation. The error components of orbital data products also tend to interact nonlinearly with hydrologic modeling uncertainty. Keeping these in mind, the present thesis fosters the development of uncertainty analysis using instantaneous satellite orbital data products (latest version 7 of 1B11, 2A25, 2A23, 2B31, 2A12) derived from the passive and active microwave sensors onboard TRMM satellite, namely TRMM Microwave Imager (TMI) and precipitation radar (PR). The study utilizes 11 years of orbital data from 2002 to 2012 over the Indian subcontinent and examines the influence of various error sources on the convective and stratiform precipitation types. Two approaches are taken up to examine uncertainty. While the first approach analyses independent contribution of error from these orbital data products, the second approach examines their combined effect. Based on the first approach, analysis conducted over the land regions of Mahanadi basin, India investigates three sources of uncertainty in detail. These include 1) errors due to improper delineation of rainfall signature within microwave footprint (rain/no rain classification), 2) uncertainty offered by the transfer function linking rainfall with TMI low frequency channels and 3) sampling errors owing to the narrow swath and infrequent visits of TRMM sensors. The second approach is hinged on evaluating the performance of rainfall estimates from each of these orbital data products by accumulating them within a spatial domain and using error decomposition methodologies. Microwave radiometers have taken unprecedented satellite images of earth’s weather, proving to be a valuable tool for quantitative estimation of precipitation from space. However, as mentioned earlier, with the widespread acceptance of microwave based precipitation products, it has also been recognized that they contain large uncertainties. One such source of uncertainty is contributed by improper detection of rainfall signature within radiometer footprints. To date, the most-advanced passive microwave retrieval algorithms make use of databases constructed by cloud or numerical weather model simulations that associate calculated microwave brightness temperature to physically plausible sample rain events. Delineation of rainfall signature from microwave footprints, also known as rain/norain classification (RNC) is an essential step without which the succeeding retrieval technique (using the database) gets corrupted easily. Although tremendous advances have been made to catapult RNC algorithms from simple empirical relations formulated for computational expedience to elaborate computer intensive schemes which effectively discriminate rainfall, a number of challenges remain to be addressed. Most of the algorithms that are globally developed for land, ocean and coastal regions may not perform well for regional catchments of small areal extent. Motivated by this fact, the present work develops a regional rainfall detection algorithm based on scattering index methodology for the land regions of study area. Performance evaluation of this algorithm, developed using low frequency channels (of 19 GHz, 22 GHz), are statistically tested for individual case study events during 2011 and 2012 Indian summer monsoonal months. Contingency table statistics and performance diagram show superior performance of the algorithm for land regions of the study region with accurate rain detection observed in 95% of the case studies. However, an important limitation of this approach is comparatively poor detection of low intensity stratiform rainfall. The second source of uncertainty which is addressed by the present thesis, involves prediction of overland rainfall using TMI low frequency channels. Land, being a radiometrically warm and highly variable background, offers a myriad of complications for overland rain retrieval using microwave radiometer (like TMI). Hence, land rainfall algorithms of TRMM TMI have traditionally incorporated empirical relations of microwave brightness temperature (Tb) with rain rate, rather than relying on physically based radiative transfer modeling of rainfall (as implemented in TMI ocean algorithm). In the present study, sensitivity analysis is conducted using spearman rank correlation coefficient as the indicator, to estimate the best combination of TMI low frequency channels that are highly sensitive to near surface rainfall rate (NSR) from PR. Results indicate that, the TMI channel combinations not only contain information about rainfall wherein liquid water drops are the dominant hydrometeors, but also aids in surface noise reduction over a predominantly vegetative land surface background. Further, the variations of rainfall signature in these channel combinations were seldom assessed properly due to their inherent uncertainties and highly non linear relationship with rainfall. Copula theory is a powerful tool to characterize dependency between complex hydrological variables as well as aid in uncertainty modeling by ensemble generation. Hence, this work proposes a regional model using Archimedean copulas, to study dependency of TMI channel combinations with respect to precipitation, over the land regions of Mahanadi basin, India, using version 7 orbital data from TMI and PR. Studies conducted for different rainfall regimes over the study area show suitability of Clayton and Gumbel copula for modeling convective and stratiform rainfall types for majority of the intraseasonal months. Further, large ensembles of TMI Tb (from the highly sensitive TMI channel combination) were generated conditional on various quantiles (25th, 50th, 75th, 95th) of both convective and stratiform rainfall types. Comparatively greater ambiguity was observed in modeling extreme values of convective rain type. Finally, the efficiency of the proposed model was tested by comparing the results with traditionally employed linear and quadratic models. Results reveal superior performance of the proposed copula based technique. Another persistent source of uncertainty inherent in low earth orbiting satellites like TRMM arise due to sampling errors of non negligible proportions owing to the narrow swath of satellite sensors coupled with a lack of continuous coverage due to infrequent satellite visits. This study investigates sampling uncertainty of seasonal rainfall estimates from PR, based on 11 years of PR 2A25 data product over the Indian subcontinent. A statistical bootstrap technique is employed to estimate the relative sampling errors using the PR data themselves. Results verify power law scaling characteristics of relative sampling errors with respect to space time scale of measurement. Sampling uncertainty estimates for mean seasonal rainfall was found to exhibit seasonal variations. To give a practical demonstration of the implications of bootstrap technique, PR relative sampling errors over the sub tropical river basin of Mahanadi, India were examined. Results revealed that bootstrap technique incurred relative sampling errors of <30% (for 20 grid), <35% (for 10 grid), <40% (for 0.50 grid) and <50% (for 0.250 grid). With respect to rainfall type, overall sampling uncertainty was found to be dominated by sampling uncertainty due to stratiform rainfall over the basin. In order to study the effect of sampling type on relative sampling uncertainty, the study compares the resulting error estimates with those obtained from latin hypercube sampling. Based on this study, it may be concluded that bootstrap approach can be successfully used for ascertaining relative sampling errors offered by TRMM-like satellites over gauged or ungauged basins lacking in in-situ validation data. One of the important goals of TRMM Ground Validation Program has been to estimate the random and systematic uncertainty associated with TRMM rainfall estimates. Disentangling uncertainty in seasonal rainfall offered by independent observations of TMI and PR enables to identify errors and inconsistencies in the measurements by these instruments. Motivated by this thought, the present work examines the spatial error structure of daily precipitation derived from the version 7 TRMM instantaneous orbital data products through comparison with the APHRODITE data over a subtropical region namely Mahanadi river basin of the Indian subcontinent for the seasonal rainfall of 6 years from June 2002 to September 2007. The instantaneous products examined include TMI and PR data products of 2A12, 2A25 and 2B31 (combined data from PR and TMI). The spatial distribution of uncertainty from these data products was quantified based on the performance metrics derived from the contingency table. For the seasonal daily precipitation over 10x10 grids, the data product of 2A12 showed greater skill in detecting and quantifying the volume of rainfall when compared with 2A25 and 2B31 data products. Error characterization using various error models revealed that random errors from multiplicative error models were homoscedastic and that they better represented rainfall estimates from 2A12 algorithm. Error decomposition technique, performed to disentangle systematic and random errors, testified that the multiplicative error model representing rainfall from 2A12 algorithm, successfully estimated a greater percentage of systematic error than 2A25 or 2B31 algorithms. Results indicate that even though the radiometer derived 2A12 is known to suffer from many sources of uncertainties, spatial and temporal analysis over the case study region testifies that the 2A12 rainfall estimates are in a very good agreement with the reference estimates for the data period considered. These findings clearly document that proper characterization of error structure offered by TMI and PR has wider implications in decision making, prior to incorporating the resulting orbital products for basin scale hydrologic modeling. The current missions of GPM envision a constellation of microwave sensors that can provide instantaneous products with a relatively negligible sampling error at daily or higher time scales. This study due to its simplicity and physical approach offers the ideal basis for future improvements in uncertainty modeling in precipitation.

碩士
中原大學
土木工程研究所
94
There are debris flow disasters in recently years in Taiwan. Since present warning system is hard to reach satisfied results, debris flows frequently cause serious loss on not only human’s life but also their property. Heavy rain is one of the factors that caused debris flows. The precipitation usually concentrates during typhoon season from May to October every year. A question as to how to estimate typhoon rainfall rapidly and accurately has become very important for early warning of debris flows. The purpose of this study is to learn about estimating typhoon rainfall by using Artificial Neural Network (ANN) with cloud temperature. The Shihmen Reservoir and its watershed are taken as example area of study,and data of cloud temperature and rainfall of invading typhoon from 1996 to 2003 are collected in this study. A temperature-rainfall model is established to predict rainfall at 3 hours later in Shihmen Reservoir. Two results are found: Firstly, rain stations with similar geographic properties have similar temperature-rainfall models. It means that landform affects rainfall condition. The past references also demonstrate this. Secondly, for the same rain station, the cloud temperature relates highly with heavy rainfall. The results display that the model performed well especially in the big typhoon events. Although for the small typhoon events the model did not perform as good as for the big ones, the errors are still acceptable. The results of this paper could serve as a fine reference for predicting debris flow induced by typhoon invasion.