International Telemetering Conference Proceedings / October 30-November 02, 1995 / Riviera Hotel, Las Vegas, Nevada / A radar scatterometer transmits a series of RF pulses and measures the total-power (energy) of the backscattered signal. Measurements of the backscattered energy from the ocean's surface can be used to infer the near-surface wind vector [7]. Accurate backscatter energy measurements are required to insure accurate wind estimates. Unfortunately, the signal measurement is noisy so a separate measurement of the noise-only total-power is subtracted from the signal measurement to estimate the echo signal energy. A common metric for evaluating the accuracy of the scatterometer energy measurement is the normalized signal variance, termed K(p). In designing a scatterometer tradeoffs in design parameters are made to minimize K(p). Spaceborne scatterometers have traditionally been based on fan-beam antennas and CW modulation for which expressions for K(p) exist. Advanced pencil-beam scatterometers, such as SeaWinds currently being developed by NASA use modulated Signals so that new K(p) expressions are required. This paper outlines the derivation of the generalized K(p) expression. While very complicated in its exact form, with a simplified geometry the K(p) expression can be related to the radar ambiguity function. The resulting analysis yields insights into the tradeoffs inherent in a scatterometer design and permits analytic tradeoffs in system performance.

International Telemetering Conference Proceedings / October 25-28, 1993 / Riviera Hotel and Convention Center, Las Vegas, Nevada / Spaceborne wind scatterometers are designed principally to measure radar backscatter from the ocean's surface for the determination of the near-surface wind direction and speed. Although measurements of the radar backscatter are made over land, application of these measurements has been limited primarily to the calibration of the instrument due to their low resolution (typically 50 km). However, a recently developed resolution enhancement technique can be applied to the measurements to produced medium-scale radar backscatter images of the earth's surface. Such images have proven useful in the study of tropical vegetation3 as well as glacial5 and sea6 ice. The technique has been successfully applied2 to Seasat scatterometer (SASS) data to achieve image resolution as fine as 3-4 km. The method can also be applied to ERS-l scatterometer data. Unfortunately, the instrument processing method employed by SASS limits the ultimate resolution which can be obtained with the method. To achieve the desired measurement overlap, multiple satellite passes are required. However, with minor modifications to future Doppler scatterometer systems (such as the NASA scatterometer [NSCAT] and its follow-on EoS-era scatterometer NEXSCAT) imaging resolutions down to 1-2 km for land/ice and 5-10 km for wind measurement may be achieved on a single pass with a moderate increase in downlink bandwidth (from 3.1 kbps to 750 kbps). This paper describes these modifications and briefly describes some of the applications of this medium-scale Ku-band imagery for vegetation studies, hydrology, sea ice mapping, and the study of mesoscale winds.

RapidScat is a Ku-band radar that measures the normalized backscatter coefficient σ0 of the Earth's surface. Launched in 2015, it currently operates on the International Space Station. Nearly one year into its mission, RapidScat measurements began exhibiting strange behavior that is believed to be caused by a change in receiver gain. Changes in gain are compensated for during post-processing, but the measurements have a lower signal-to-noise ratio (SNR). Calibration and validation of σ0 measurements from this low SNR state are performed using extended land targets with various signal strengths. Study areas include the Amazon rainforest, Congo rainforest, Argentina pampas, two regions in the Sahara desert, and a desert region in Australia. The effects of seasonal, azimuthal, incidence angle and local-time-of-day variations on σ0 are studied using data from two Ku-band sensors, QuikSCAT and RapidScat, for each study area. Calibration is performed comparing RapidScat data from all SNR states to QuikSCAT data as well as comparing RapidScat low SNR state data to the nominal (high SNR) state data. Results from both calibrations are consistent with each other. Results suggest that σ0 is unbiased by noise for the ranges of σ0 covered in this study (-7 dB to -27 dB). However, the second low SNR state vertically polarized σ0 appears to be biased lower than would be expected from year-to-year seasonal variation. The third low SNR state σ0 appears unbiased compared to the nominal RapidScat SNR σ0.

No description available.

L’industrie des nanotechnologies est un monde en constante évolution. Les améliorations dans les techniques de fabrication permettent de définir des composants de plus en plus petits. Afin de vérifier les dimensions fabriquées, la métrologie doit s’adapter et être capable de fournir des analyses basées sur des mesures optiques fiables et répétables. Ces travaux se focalisent dans un premier temps sur les procédés de fabrication des échantillons. Les techniques de dépôts, de photolithographies et de gravure sont présentées. Ces techniques nécessitent des outils de métrologie adaptés permettant un contrôle en ligne. Les équipements de métrologie disponibles sont donc présentés en se focalisant sur les techniques par imagerie comme la microscopie électronique à balayage ou transmission et les techniques par inversion telle que l’ellipsométrie et la scatterometrie. Le troisième chapitre est dédié aux applications de ces techniques en production. Les empilements étudiés sont généralement constitués de couches innovantes. La dernière partie est axée sur des méthodes de caractérisation de la rugosité par diffusion lumineuse sur des réseaux périodiques. / Nanotechnology industry is a world in constant evolution. Improvements in manufacturing techniques are used to define smaller and smaller components. To verify dimensions, metrology has to be improved to be able to provide reliable and repeatable analysis. This work focuses first on manufacturing process samples. Deposition techniques, photolithography and etching are introduced. These techniques require metrology tools adapted for in-line monitoring. Metrology equipments introduced in this thesis highlight the application of imaging techniques such as scanning electron microscopy or transmission and inversion techniques such as ellipsometry and scatterometry. The third chapter is dedicated to the application of these techniques to monitoring production. Thin films inspected are generally innovative layers. The last part is focused on methods for roughness characterization by light scattering on periodic gratings.

Radar observations of the ocean surface are becoming increasingly important. Common applications are wind retrieval and global weather forecasting and characterization. Because of the common use of ocean radar measurements, it is important to understand the sensitivity of the backscatter to both radar parameters and surface parameters. At near-normal incidence angles, it has been assumed that the radar backscatter exhibits little or no azimuth dependence (Colton, 1989). However, recent data taken by the BYU YSCAT radar system suggests that this is not the case. At an incidence angle of 10°, the YSCAT radar data shows from a fraction of a decibel to up to 10 decibels of azimuth modulation, depending on the surface conditions. In this thesis, a physical optics approach is used with a two-dimensional surface model to derive the electromagnetic backscatter from the ocean surface. If the waves on the ocean surface are directed, azimuth modulation is predicted at near-normal incidence angles. The effects of surface and radar parameters on the azimuth modulation are studied, and the results are compared to data taken by the YSCAT radar system. It is shown that the theory correctly predicts of the shape of the curve when the normalized radar cross-section is plotted as a function of azimuth angle. The theory also predicts the correct trend of the modulation magnitude as function the surface roughness. However, the simplifications in the model limit its prediction of the frequency dependence of the modulation. Relaxing some of the assumptions of the model is likely to correct this problem.

Microwave scatterometers, which use radar backscatter measurements to infer the near-surface wind vector, are unique in their ability to monitor global wind vectors at high resolutions. However, scatterometer observations which are contaminated by land proximity or rain events produce wind estimates which have increased bias and variability, making them unreliable for many applications. Fortunately, the effects of these sources of contamination can be mitigated. Land contamination of backscatter measurements occurs when land partially fills the antenna illumination area. This reduces and masks the wind-induced backscatter signal. Land contamination is mitigated by quantifying the amount of contamination in a single observation using a metric referred to as the land contribution ratio (LCR). LCR levels which give rise to inadmissible levels of error in the wind estimates are determined and used to discard land-contaminated observations. Using this method results in contamination-free wind estimates which can be made as close to the coast as 5 km, an improvement of 25 km compared to previous methods. Rain contamination of scatterometer observations results from rain-induced scattering effects which modify the wind-induced backscatter. Rain backscatter effects are modeled phenomenologically to assess the impact of rain on the observed backscatter. Given the backscatter effects of wind and rain, there are three estimators: wind-only (WO), simultaneous wind and rain (SWR) and rain-only (RO), which have optimal performance in different wind and rain conditions. Rain contamination of wind estimates is mitigated using a new Bayes estimator selection (BES) technique which optimally selects WO, SWR, or RO estimates as they are most appropriate. BES is a novel adaptation of Bayes decision theory to operate on parameter estimates which may have different dimensions. The BES concept is extended to include prior selection and noise reduction techniques which generalizes BES to a wider variety of wind fields and further increase wind estimate accuracy. Overall, BES has wind estimation performance which surpasses that of either the WO or SWR wind estimates individually, and also provides a viable rain-impact flag.

La miniaturisation des composants impose à l’industrie de la micro-électronique de trouver des techniques de caractérisation fiables rapides et si possible à moindre coût. Les méthodes optiques telles que la scattérométrie se présentent aujourd’hui comme des alternatives prometteuses répondant à cette problématique de caractérisation. Toutefois, l’ensemble des méthodes scattérométriques nécessitent un certain nombre d’hypothèses pour assurer la résolution d’un problème inverse et notamment la connaissance de la forme géométrique de la structure à tester. Le modèle de structure supposé conditionne la qualité même de la caractérisation. Dans cette thèse, nous proposons l’utilisation des réseaux de neurones comme outils d’aide à la décision en amont de toute méthode de caractérisation. Nous avons validé l’utilisation des réseaux de neurones dans le cadre de la reconnaissance des formes géométriques de l’échantillon à tester par la signature optique utilisée dans toute étape de caractérisation scattérométrique. Tout d’abord, le cas d’un défaut lithographique particulier lié à la présence d’une couche résiduelle de résine au fond des sillons est étudié. Ensuite, nous effectuons une analyse de détection de défaut de modèle utilisé dans la résolution du problème inverse. Enfin nous relatons les résultats obtenus dans le cadre de la sélection de modèles géométriques par réseaux de neurones en amont d’un processus classique de caractérisation scattérométrique. Ce travail de thèse a montré que les réseaux de neurones peuvent bien répondre à la problématique de classification en scattérométrie ellipsométrique et que l’utilisation de ces derniers peut améliorer cette technique optique de caractérisation / The miniaturization of components in the micro-electronics industry involves the need of fast reliable technique of characterization with lower cost. Optical methods such as scatterometry are today promising alternative to this technological need. However, scatterometric method requires a certain number of hypothesis to ensure the resolution of an inverse problem, in particular the knowledge of the geometrical shape of the structure under test. The assumed model of the structure determines the quality of the characterization. In this thesis, we propose the use of neural networks as decision-making tools upstream of any characterization method. We validated the use of neural networks in the context of recognition of the geometrical shapes of the sample under testing by the use of optical signature in any scatterometric characterization process. First, the case of lithographic defect due to the presence of a resist residual layer at the bottom of the grooves is studied. Then, we carry out an analysis of model defect in the inverse problem resolution. Finally, we report results in the context of selection of geometric models by neural networks upstream of a classical scatterometric characterization process. This thesis has demonstrated that neural networks can well answer the problem of classification in ellipsometric scatterometry and their use can improve this optical characterization technique

Les dimensions critiques des circuits intégrés diminuent continuellement au coursdes ans selon la loi de Moore. Les valeurs typiques sont aujourd’hui de 28nm,et seront de 22nm dans 18 mois. La photo-lithographie optique demeure encore latechnique la plus économique pour la production de masse. L’ouverture numériquedes objectifs atteint 1.30, grâce à l’introduction d’eau entre la lentille et la plaquette.La conséquence directe de ces grandes ouvertures est la réduction de la profondeurde champs de l’ordre d’une centaine de nanomètres. Le procédé photo-lithographiqueperdant de la latitude, le contrôle dimensionnel intra-cellule devient une nécessité.La variation dimensionnelle provient au premier ordre à la fois de la variation duplan focal dans le champ image, et aussi de la non uniformité de l’illumination duréticule. Pour contrôler cette variation, une boucle de régulation a été mise en placepour ajuster uniquement l’énergie des lots de production. On corrige ainsi de fait unmauvais focus par une compensation en énergie.Pour ne pas altérer l’image dans la résine, il est important de pouvoir dissocierles deux effets et adresser les causes de dégradation de l’image séparément. Le sujetde cette thèse est précisément de trouver un moyen de décorréler les deux paramètresaffectant l’uniformité de la dimension critique. L’idée principale est de trouver à lafois les motifs et les modèles théoriques pouvant conduire à discriminer des imagesselon leur sensibilité, soit au focus, soit à la dose / Following the ITRS roadmap, the critical dimension of the circuits are continuouslynarrowing. Optical Lithography still remains the cheapest way forintegrated circuits mass production. If the resists properties and the exposure wavelengthreduction had an important contribution to this result, the lens numericalaperture increase had a decisive impact. The numerical aperture is currently reaching1,30 thanks to the usage of water as immersion fluid between the lens andthe wafer. Future lens are targeting in a near future a 1,70 numerical aperture withimmersion fluids at higher refractive index. A direct consequence of these wider numericalaperture’s is the reduction of the depth of focus to few tens of nanometers,reducing the process windows and then the integrated circuits manufacturability. Inaddition the pure numerical aperture effect on focus, off axis illumination is leadingto amplify the reticle critical dimension variations, and the intrafield focus controlbecomes more and more crucial.The last scanner generation provides some tools to adjust the intrafield energy.As the two effects appear to compensate each of them critical dimension variation,it becomes very important to be able to dissociate the effect of one from the otherin order to select the most efficient mean to get the greater process windows. Moreover, the average value compensations must feed accurately the Run to Run feedback loop for the next exposed wafers.The purpose of this thesis is to find a way to un-correlate the various parametersaffecting the critical dimension uniformity. Some researchers tried to design specificfeatures whose shape modifications due to focus offset could be detected as an overlayerror measured by the appropriate tool, but the limitations seem to be actuallyreached. New tools, as scatterometers, could provide a more precise information.The desire output of this thesis would be to provide a methodology to allow an inline intrafield focus follow up for the future technologies at 20nm half pitch

Two surface-based, portable, S-band and C-band polarimetric scatterometers intended for in situ measurements of both terrain and the ocean’s surface are presented. The scatterometers' layout, hardware design, measurement accuracy, calibration, and signal processing concepts are described. To augment in situ geophysical observations, researchers have often employed in situ scatterometers for validating satellite-based retrievals and also for their innate ability to monitor geophysical variations of localized regions with fine temporal resolution. Backscatter measurement variability due to system effects is presented, providing the fundamental basis for the quantitative analysis of data. Sample polarimetric retrievals are presented for asphalt pavement and grass.