Engineering PapersSearch

Engineering topics

Stanley, Thomas

Publications and source records attributed to Stanley, Thomas.

Landslide Mapping Along the Karnali Highway, Nepal using High-Resolution Imagery

The Karnali highway (Figure 1) is the only major transportation link that connects the remote Karnali region to the provincial capital in Province 6 of Nepal. This area becomes inaccessible by roads during every rainy season due to landslides. Despite the known landslide frequency, there have been no systematic landslide inventories conducted along this highway to date. Recent advancements in remote-sensing technologies have significantly increased our ability to map landslides of various sizes rapidly with less in situ surveys or human interaction. Landslide susceptibility, hazard and risk studies require a complete landslide inventory, which might only be possible from very high-resolution (VHR) and high-resolution (HR) imagery. Recent launch of Sentinel-2 in 2015 has provided free access to HR imagery enabling landslide detection at finer scales then what was possible with previous open source satellite imagery obtained from Landsat and ASTER. Satellites providing VHR imagery are commercially owned, expensive and not freely available expect for when disasters charter is activated. NextView licensing agreement, a partnership between the US government and US commercial vendors provides access to VHR imagery to federal agencies in support of scientific research [1]. This partnership provides access to VHR imagery obtained from the DigitalGlobe (DG) constellation which enables mapping of small landslides (< 100 m2). In this study, VHR imagery from DG and HR imagery from Sentinel-2 will be used to map landslides along the Karnali highway using a semi automatic method based on object-oriented analysis (OOA) to create most recent and up-to-date landslide inventory. The effectiveness of this remote sensing based landslide inventory to produce a susceptibility map and its predictive capacity will be tested.

landslide

Effects of Inventory Bias on Landslide Susceptibility Calculations

Many landslide inventories are known to be biased, especially inventories for large regions such as Oregons SLIDO or NASAs Global Landslide Catalog. These biases must affect the results of empirically derived susceptibility models to some degree. We evaluated the strength of the susceptibility model distortion from postulated biases by truncating an unbiased inventory. We generated a synthetic inventory from an existing landslide susceptibility map of Oregon, then removed landslides from this inventory to simulate the effects of reporting biases likely to affect inventories in this region, namely population and infrastructure effects. Logistic regression models were fitted to the modified inventories. Then the process of biasing a susceptibility model was repeated with SLIDO data. We evaluated each susceptibility model with qualitative and quantitative methods. Results suggest that the effects of landslide inventory bias on empirical models should not be ignored, even if those models are, in some cases, useful. We suggest fitting models in well-documented areas and extrapolating across the study region as a possible approach to modelling landslide susceptibility with heavily biased inventories.

landslide

Rainfall-Triggered Landslides Bury Sri Lankan Villages

On the afternoon of May 17th, 2016, a major landslide event caused at least 92 deaths, with 109 still missing*. The site was rated highly susceptible to landslides in a new global landslide susceptibility map. GPM precipitation data suggest that both antecedent and current rainfall as well as complex topography played a role in the slope failures.

IMERG

Software Suite to Support In-Flight Characterization of Remote Sensing Systems

A characterization software suite was developed to facilitate NASA's in-flight characterization of commercial remote sensing systems. Characterization of aerial and satellite systems requires knowledge of ground characteristics, or ground truth. This information is typically obtained with instruments taking measurements prior to or during a remote sensing system overpass. Acquired ground-truth data, which can consist of hundreds of measurements with different data formats, must be processed before it can be used in the characterization. Accurate in-flight characterization of remote sensing systems relies on multiple field data acquisitions that are efficiently processed, with minimal error. To address the need for timely, reproducible ground-truth data, a characterization software suite was developed to automate the data processing methods. The characterization software suite is engineering code, requiring some prior knowledge and expertise to run. The suite consists of component scripts for each of the three main in-flight characterization types: radiometric, geometric, and spatial. The component scripts for the radiometric characterization operate primarily by reading the raw data acquired by the field instruments, combining it with other applicable information, and then reducing it to a format that is appropriate for input into MODTRAN (MODerate resolution atmospheric TRANsmission), an Air Force Research Laboratory-developed radiative transport code used to predict at-sensor measurements. The geometric scripts operate by comparing identified target locations from the remote sensing image to known target locations, producing circular error statistics defined by the Federal Geographic Data Committee Standards. The spatial scripts analyze a target edge within the image, and produce estimates of Relative Edge Response and the value of the Modulation Transfer Function at the Nyquist frequency. The software suite enables rapid, efficient, automated processing of ground truth data, which has been used to provide reproducible characterizations on a number of commercial remote sensing systems. Overall, this characterization software suite improves the reliability of ground-truth data processing techniques that are required for remote sensing system in-flight characterizations.

Stanley, Thomas

Simplified Vicarious Radiometric Calibration

A measurement-based radiance estimation approach for vicarious radiometric calibration of spaceborne multispectral remote sensing systems has been developed. This simplified process eliminates the use of radiative transfer codes and reduces the number of atmospheric assumptions required to perform sensor calibrations. Like prior approaches, the simplified method involves the collection of ground truth data coincident with the overpass of the remote sensing system being calibrated, but this approach differs from the prior techniques in both the nature of the data collected and the manner in which the data are processed. In traditional vicarious radiometric calibration, ground truth data are gathered using ground-viewing spectroradiometers and one or more sun photometer( s), among other instruments, located at a ground target area. The measured data from the ground-based instruments are used in radiative transfer models to estimate the top-of-atmosphere (TOA) target radiances at the time of satellite overpass. These TOA radiances are compared with the satellite sensor readings to radiometrically calibrate the sensor. Traditional vicarious radiometric calibration methods require that an atmospheric model be defined such that the ground-based observations of solar transmission and diffuse-to-global ratios are in close agreement with the radiative transfer code estimation of these parameters. This process is labor-intensive and complex, and can be prone to errors. The errors can be compounded because of approximations in the model and inaccurate assumptions about the radiative coupling between the atmosphere and the terrain. The errors can increase the uncertainty of the TOA radiance estimates used to perform the radiometric calibration. In comparison, the simplified approach does not use atmospheric radiative transfer models and involves fewer assumptions concerning the radiative transfer properties of the atmosphere. This new technique uses two neighboring uniform ground target areas having different reflectance values. The target areas can be natural or artificial and must be large enough to minimize adjacent-pixel contamination effects. The radiative coupling between the atmosphere and the terrain needs to be approximately the same for the two targets. This condition can be met for relatively uniform backgrounds when the distance between the targets is within a few hundred meters. For each target area, the radiance leaving the ground in the direction of the satellite is measured with a radiometrically calibrated spectroradiometer. Using the radiance measurements from the two targets, atmospheric adjacency and atmospheric scattering effects can be subtracted, thereby eliminating many assumptions about the atmosphere and the radiative interaction between the atmosphere and the terrain. In addition, the radiometrically calibrated spectroradiometer can be used with a known reflectance target to estimate atmospheric transmission and diffuse- to-global ratios without the need for ancillary sun photometers. Several comparisons between the simplified method and traditional techniques were found to agree within a few percent. Hence, the simplified method reduces the overall complexity of performing vicarious calibrations and can serve as a method for validating traditional radiative transfer models

Stanley, Thomas

A Non-Radiative Transfer Approach to Radiometric Vicarious Calibration

TOA (top-of-atmosphere) radiance from high-spatial-resolution satellite imagery systems is important for a wide variety of research and applications. Many research initiatives require data with absolute radiometric accuracy better than a few percent. The conversion of satellite digital numbers to radiance depends on accurate radiometric calibration. A common method for determining and validating radiometric calibrations is to rely upon vicarious calibration approaches. Historically, vicarious calibration methods use radiative transfer codes with ground-based atmosphere and surface reflectance or radiance inputs for estimating TOA radiance values. These TOA radiance values are compared against the satellite digital numbers to determine the radiometric calibration. However, the radiative transfer codes used depend on many assumptions about the aerosol properties and the atmospheric point spread function. A measurement-based atmospheric radiance estimation approach for high-spatial-resolution, multispectral, visible/near-infrared sensors is presented that eliminates the use of radiative transfer codes and many of the underlying assumptions. A comparison between the radiative transfer and non-radiative transfer approaches is made.

Ryan, Robert

Proceedings of the 2006 Civil Commercial Imagery Evaluation Workshop

The Joint Agency Commercial Imagery Evaluation (JACIE) team is a collaborative interagency working group formed to leverage different government agencies' capabilities for the characterization of commercial remote sensing products. The team is composed of staff from the National Aeronautics and Space Administration (NASA), the National Geospatial-Intelligence Agency (NGA), and the U.S. Geological Survey (USGS). Each JACIE agency has a vested interest in the purchase and use of commercial imagery to support government research and operational applications. The intent of the 2006 workshop is to exchange information regarding the characterization and application of commercial imagery used by the government. The main focus of previous workshops has been on high-resolution satellite imagery from systems; such as, IKONOS (Space Imaging, Inc.), QuickBird (DigitalGlobe, Inc.), and OrbView-3 (ORBIMAGE). This workshop is being expanded to cover all civil medium- and high-resolution commercial imagery used by the government.

Stanley, Thomas

Emerging Techniques for Vicarious Calibration of Visible Through Short Wave Infrared Remote Sensing Systems

Autonomous Visible to SWIR ground-based vicarious Cal/Val will be an essential Cal/Val component with such a large number of systems. Radiometrically calibrated spectroradiometers can improve confidence in current ground truth data through validation of radiometric modeling and validation or replacement of traditional sun photometer measurement. They also should enable significant reduction in deployed equipment such as equipment used in traditional sun photometer approaches. Simple, field-portable, white-light LED calibration source shows promise for visible range (420-750 nm). Prototype demonstrated <0.5% drift over 10-40 C temperature range. Additional complexity (more LEDs) will be necessary for extending spectral range into the NIR and SWIR. LED long lifetimes should produce at least several hundreds of hours or more of stability, minimizing the need for expensive calibrations and supporting long-duration field campaigns.

Ryan, Robert E.

Spatial Resolution Characterization for AWiFS Multispectral Images

This viewgraph presentation describes the spatial resolution of the AWiFS multispectral images characterized by an estimation of the Modulation Transfer Function (MTF) at Nyquist frequency. The contents include: 1) MTF Analysis; 2) Target Analysis; 3) "Pulse Target"; 4) "Pulse" Method; 5) Target Images; 6) Bridge Profiles; 7) MTF Calculation; 8) MTF Results; and 9) Results Summary.

Blonski, Slawomir

Spatial Resolution Characterization for AWiFS Multispectral Images

Within the framework of the Joint Agency Commercial Imagery Evaluation program, the National Aeronautics and Space Administration, the National Geospatial-Intelligence Agency, and the U.S. Geological Survey cooperate in the characterization of high-to-moderate-resolution commercial imagery of mutual interest. One of the systems involved in this effort is the Advanced Wide Field Sensor (AWiFS) onboard the Indian Remote Sensing (IRS) Reourcesat-1 satellite, IRS-P6. Spatial resolution of the AWiFS multispectral images was characterized by estimating the value of the system Modulation Transfer Function (MTF) at the Nyquist spatial frequency. The Nyquist frequency is defined as half the sampling frequency, and the sampling frequency is equal to the inverse of the ground sample distance. The MTF was calculated as a ratio of the Fourier transform of a profile across an AWiFS image of the Lake Pontchartrain Causeway Bridge and the Fourier transform of a profile across an idealized model of the bridge for each spectral band evaluated. The mean MTF value for the AWiFS imagery evaluated was estimated to be 0.1.

Blonski, Slawomir

Spatial Resolution Characterization for Aerial Digital Imagery

The U.S. Geological Survey (USGS) and the National Aeronautics and Space Administration have jointly conducted characterizations of aerial digital imagery from commercial data providers. In a typical scenario, a vendor acquired imagery over the Stennis Space Center test range and provided a common data package to SSC. SSC personnel analyzed the geopositional accuracy and spatial resolution of the images, then documented the characterization results in a report and delivered the report to the USGS. Spatial resolution characterization was based on edge response measurements using one of the SSC edge targets and the tilted-edge technique. Relative Edge Response (RER) has been estimated for aerial digital imagery from several commercial data providers. RER is one of the engineering parameters used in the General Image Quality Equation to provide predictions of imaging system performance expressed in terms of the National Imagery Interpretability Rating Scale.

Blonski, Slawomir

Edge Response and NIIRS Estimates for Commercial Remote Sensing Satellites

Spatial resolution of panchromatic imagery from commercial remote sensing satellites was characterized based on edge response measurements using edge targets and the tilted-edge technique. Relative Edge Response (RER) was estimated as a geometric mean of normalized edge response differences measured in two directions of image pixels at points distanced from the edge by -0.5 and 0.5 of ground sample distance. RER is one of the engineering parameters used in the General Image Quality Equation to provide predictions of imaging system performance expressed in terms of the National Imagery Interpretability Rating Scale (NIIRS). By assuming a plausible range of signal-to-noise ratio and assessing the effects of Modulation Transfer Function compensation, the NIIRS estimates were made and then compared with vendor-provided values and evaluations conducted by the National Geospatial-Intelligence Agency.

Blonski, Slawomir

Atmospheric Correction of High-Spatial-Resolution Commercial Satellite Imagery Products Using MODIS Atmospheric Products

Remotely sensed ground reflectance is the basis for many inter-sensor interoperability or change detection techniques. Satellite inter-comparisons and accurate vegetation indices such as the Normalized Difference Vegetation Index, which is used to describe or to imply a wide variety of biophysical parameters and is defined in terms of near-infrared and redband reflectance, require the generation of accurate reflectance maps. This generation relies upon the removal of solar illumination, satellite geometry, and atmospheric effects and is generally referred to as atmospheric correction. Atmospheric correction of remotely sensed imagery to ground reflectance, however, has been widely applied to only a few systems. In this study, we atmospherically corrected commercially available, high spatial resolution IKONOS and QuickBird imagery using several methods to determine the accuracy of the resulting reflectance maps. We used extensive ground measurement datasets for nine IKONOS and QuickBird scenes acquired over a two-year period to establish reflectance map accuracies. A correction approach using atmospheric products derived from Moderate Resolution Imaging Spectrometer data created excellent reflectance maps and demonstrated a reliable, effective method for reflectance map generation.

Pagnutti, Mary

Atmospheric Correction of High-Spatial-Resolution Commercial Satellite Imagery Products Using MODIS Atmospheric Products

Remotely sensed ground reflectance is the foundation of any interoperability or change detection technique. Satellite intercomparisons and accurate vegetation indices, such as the Normalized Difference Vegetation Index (NDVI), require the generation of accurate reflectance maps (NDVI is used to describe or infer a wide variety of biophysical parameters and is defined in terms of near-infrared (NIR) and red band reflectances). Accurate reflectance-map generation from satellite imagery relies on the removal of solar and satellite geometry and of atmospheric effects and is generally referred to as atmospheric correction. Atmospheric correction of remotely sensed imagery to ground reflectance has been widely applied to a few systems only. The ability to obtain atmospherically corrected imagery and products from various satellites is essential to enable widescale use of remotely sensed, multitemporal imagery for a variety of applications. An atmospheric correction approach derived from the Moderate Resolution Imaging Spectroradiometer (MODIS) that can be applied to high-spatial-resolution satellite imagery under many conditions was evaluated to demonstrate a reliable, effective reflectance map generation method. Additional information is included in the original extended abstract.

Pagnutti, Mary

The ASPRS Digital Imagery Product Guideline Project

The American Society for Photogrammetry and Remote Sensing (ASPRS) Primary Data Acquisition Division is developing a Digital Imagery Product Guideline in conjunction with NASA, the U.S. Geological Survey (USGS), the National Imagery and Mapping Agency (NIMA), academia, and industry. The goal of the guideline is to offer providers and users of digital imagery a set of recommendatons analogous those defined by the ASPRS Aerial Photography 1995 Draft Standard for film-based imagery. This article offers a general outline and description of the Digital Imagery Product Guideline and Digital Imagery Tutorial/Reference documents for defining digital imagery requirements.

Ryan, Robert