Browsing by Author "Shao, Yang"

Now showing 1 - 20 of 66
  • Thumbnail Image
    2012 OGIS Symposium Program
    McGee, John; Wynne, Randolph H.; Shao, Yang (2012-04-13)
    List of all proceedings from the Virginia Tech GIS Symposium, held on April 13, 2012.
  • Thumbnail Image
    2013 OGIS Symposium Program
    McGee, John; Wynne, Randolph H.; Shao, Yang (2013-04-19)
    List of all proceedings from the Virginia Tech GIS Symposium, held on April 19, 2013.
  • Thumbnail Image
    A 4700-Year Record of Lake Evolution and Fire History for Laguna Limon, Dominican Republic
    McVay, Jason Lyle (Virginia Tech, 2013-05-23)
    Fire is a primary driver of environmental change that can originate from natural or human ignition. Macroscopic charcoal (>125 "m) deposited into lake sediment is a record of a local fire event, whereas microscopic charcoal indicates fire activity on a broad landscape scale. Patterns of charcoal deposition may shed light on both human activities and climate history over long-time scales. Whether lowland Caribbean forests have experienced natural fire regimes over the long-term is unknown. Laguna Limón is a little-studied, large, freshwater lake on the northeastern coast of the Dominican Republic. We extracted four overlapping sediment cores totaling 315 cm in depth, and conducted analysis of macroscopic charcoal (2-cm), microscopic charcoal (16-cm), and loss-on-ignition (1-cm) to examine the long-term fire and environmental history of the area. Loss-on-ignition data established that the lake has only recently become organic rich, and was likely open to the sea as a low energy bay until 1400 Cal. Yr BP. The lake existed briefly as a wetland before transitioning to the modern freshwater lake 1200 Cal. Yr BP. Macroscopic charcoal was most abundant in the freshwater section of the core while microscopic charcoal peaked near the bottom of the core, and aligns well with other regional microscopic charcoal records. Overall the charcoal record reflects a combination of climatic and anthropogenic related charcoal deposition suggesting that fire has played an active role in the environmental history Laguna Limón.
  • Thumbnail Image
    Accuracy Assessment of the NLCD 2006 Impervious Surface for Roanoke and Blacksburg
    Zhao, Suwen; Feng, Leyang; Shao, Yang; Dymond, Randel L. (2014)
    Impervious surface map products are important for the study of urbanization, urban heat island effects, watershed hydrology, water pollution, and ecosystem services in general. At the conterminous US scale, impervious surfaces are mapped for 2001 and 2006. The accuracy of the 2006 NLCD impervious surface, however, has not been thoroughly examined, especially for small and intermediate size cities (e.g., regional city). In this study, we selected two transects in two cities and visually interpreted aerial photo to develop impervious surface reference maps. We then compared percent impervious surface of the NLCD and aerial photo-interpreted reference maps. The comparison was conducted at 90m resolution to minimize the errors in image registration. Overall, we found that the 2006 NLCD impervious surface matched well with our reference data, although slight skewness at two extremes is present. The R² and RMSE statistics improved when the two datasets are compared at coarse aggregation levels (e.g. 180m).
  • Thumbnail Image
    Agricultural practices and perceptions of climate change in Keur Samba Guéye village, Senegal, West Africa
    Diaw, Adja Adama (Virginia Tech, 2013-06-11)
    This research uses a mixed methods approach to analyze recent climate and land use changes, and farmers\' perceptions of climate change and its impacts on traditional agriculture in the village of Keur Samba Guéye (KSG). This work looks at the influence of social beliefs in adoption of new strategies by small farmers in this region, a topic that has received little or no study to date. Traditional agriculture in KSG is not very productive at present because of the impoverishment of the area and traditional agricultures strong dependency on natural climatic conditions. In this research, I identified recent climatic trends, documented changes in land use/land cover (LULC) from 1989 to 2011, and assessed farmers\' perceptions of climate change and their responses to such changes. To document climate trends and LULC, I analyzed climate data of twelve meteorological stations located across the country and created a classification of satellite images of KSG for two time periods. To examine farmers\' perceptions and agricultural practices, I conducted surveys of the farmers of KSG and in surrounding villages. Most farmers reported negative impacts of climate change on their agriculture activities, and interest in adopting new agricultural strategies despite long-standing tradition. Increasing temperatures and irregularity of rainfall may have negatively impacted crop yields, but more climate data are needed to clarify this phenomenon. LULC has been influenced by both climate change and human pressure; agricultural land has declined, while bare soils have increased. Several recommendations are provided that may help farmers to cope with changing climate.
  • Thumbnail Image
    Analysis of Crop Phenology Using Time-Series MODIS Data and Climate Data
    Ren, Jie; Campbell, James B. Jr.; Shao, Yang; Thomas, R. Quinn (2014)
    Understanding crop phenology is fundamental to agricultural production, management, planning and decision-making. In the continental United States, key phenological stages are strongly influenced by meteorological and climatological conditions. This study used remote sensing satellite data and climate data to determine key phenological states of corn and soybean and evaluated estimates of these phenological parameters. A time series of Moderate Resolution Imaging Spectrometer (MODIS) Normalized Difference Vegetation Index (NDVI) 16-day composites from 2001 to 2013 was analyzed with the TIMESAT program to automatically retrieve key phenological stages such as the start of season (emergence), peak (heading) and end of season (maturity). These stages were simulated with 6 hourly temperature data from 1980 to 2013 on the basis of crop model under the Community Land Model (CLM) (version 4.5). With these two methods, planting date, heading date, harvesting date, and length of growing season from 2001 to 2013 were determined and compared. There should be a good correlation between estimates derived from satellites and estimates produced with the climate data based on the crop model.
  • Thumbnail Image
    Assessing annual urban change and its impacts on evapotranspiration
    Wan, Heng (Virginia Tech, 2020-06-19)
    Land Use Land Cover Change (LULCC) is a major component of global environmental change, which could result in huge impacts on biodiversity, water yield and quality, climate, soil condition, food security and human welfare. Of all the LULCC types, urbanization is considered to be the most impactful one. Monitoring past and current urbanization processes could provide valuable information for ecosystem services evaluation and policy-making. The National Land Cover Database (NLCD) provides land use land cover data covering the entire United States, and it is widely used as land use land cover data input in numerous environmental models. One major drawback of NLCD is that it is updated every five years, which makes it unsatisfactory for some models requiring land use land cover data with a higher temporal resolution. This dissertation integrated a rich time series of Landsat imagery and NLCD to achieve annual urban change mapping in the Washington D.C. metropolitan area by using time series data change point detection methods. Three different time series change point detection methods were tested and compared to find out the optimal one. One major limitation of using the above time series change point detection method for annual urban mapping is that it relies heavily on NLCD, thus the method is not applicable to near-real time monitoring of urban change. To achieve the near real-time urban change identification, this research applied machine learning-based classification models, including random forest and Artificial Neural Networks (ANN), to automatically detect urban changes by using a rich time series of Landsat imagery as inputs. Urban growth could result in a higher probability of flooding by reducing infiltration and evapotranspiration (ET). ET plays an important role in stormwater mitigation and flood reduction, thus assessing the changes of ET under different urban growth scenarios could yield valuable information for urban planners and policy makers. In this study, spatial-explicit annual ET data at 30-m resolution was generated for Virginia Beach by integrating daily ET data derived from METRIC model and Landsat imagery. Annual ET rates across different major land cover types were compared, and the results indicated that converting forests to urban could result in a huge deduction in ET, thus increasing flood probability. Furthermore, we developed statistical models to explain spatial ET variation using high resolution (1m) land cover data. The results showed that annual ET will increase with the increase of the canopy cover, and it would decrease with the increase of impervious cover and water table depth.
  • Thumbnail Image
    Assessing Deep Convolutional Neural Networks and Assisted Machine Perception for Urban Mapping
    Shao, Yang; Cooner, Austin J.; Walsh, Stephen J. (MDPI, 2021-04-15)
    High-spatial-resolution satellite imagery has been widely applied for detailed urban mapping. Recently, deep convolutional neural networks (DCNNs) have shown promise in certain remote sensing applications, but they are still relatively new techniques for general urban mapping. This study examines the use of two DCNNs (U-Net and VGG16) to provide an automatic schema to support high-resolution mapping of buildings, road/open built-up, and vegetation cover. Using WorldView-2 imagery as input, we first applied an established OBIA method to characterize major urban land cover classes. An OBIA-derived urban map was then divided into a training and testing region to evaluate the DCNNs’ performance. For U-Net mapping, we were particularly interested in how sample size or the number of image tiles affect mapping accuracy. U-Net generated cross-validation accuracies ranging from 40.5 to 95.2% for training sample sizes from 32 to 4096 image tiles (each tile was 256 by 256 pixels). A per-pixel accuracy assessment led to 87.8 percent overall accuracy for the testing region, suggesting U-Net’s good generalization capabilities. For the VGG16 mapping, we proposed an object-based framing paradigm that retains spatial information and assists machine perception through Gaussian blurring. Gaussian blurring was used as a pre-processing step to enhance the contrast between objects of interest and background (contextual) information. Combined with the pre-trained VGG16 and transfer learning, this analytical approach generated a 77.3 percent overall accuracy for per-object assessment. The mapping accuracy could be further improved given more robust segmentation algorithms and better quantity/quality of training samples. Our study shows significant promise for DCNN implementation for urban mapping and our approach can transfer to a number of other remote sensing applications.
  • Thumbnail Image
    Biophysical and Climate Analysis of the Mountain Pine Beetle (Dendroctonus ponderosae) Infestations in the Crown of the Continent, 1962 to 2014
    Garza, Mario Nicholas (Virginia Tech, 2017-02-22)
    Mountain pine beetles (Dendroctonus ponderosae) are native insects that have decimated millions of hectares of mature pine (Pinaceae) forests in western North America. The purpose of this study is to investigate biophysical and climatic correlates of Mountain Pine Beetle (MPB) insect outbreaks in the Crown of the Continent Ecosystem (CCE) from 1962 to 2014 using Aerial Detection Survey (ADS) and climate data. Specific objectives were: 1) to develop statistical models to determine how selected biophysical correlates (slope, aspect, elevation, and latitude) and 2) to understand how local and global climate variables relate to the extent of the MPB infestations in the CCE, and 3) to contextualize the results of the models with historical climate data. Overall, the major findings of this study are: 1) despite its limitations, the ADS data seems suitable for analysis of beetle damage with respect to climate and topographic factors, on a regional scale, 2) there appears to be a link between local biophysical factors and winter precipitation and TPA within the CCE, and 3) a combination of a negative-phase PDO and La Niña is important in forecasting a decline in MPB spread, during a given year. This study is the first, to our knowledge, to explore spatio-temporal patterns of MPB outbreaks using biophysical factors, and both local and global climate variables, over a fifty-year timespan in the CCE. In the future, additional geospatial analyses may enable a landscape assessment of factors contributing to variability of MPB infestation and damage as this insect continues to spread.
  • Thumbnail Image
    Characterizing and modeling wet stream length dynamics in Appalachian headwaters
    Jensen, Carrie Killeen (Virginia Tech, 2018-05-03)
    Headwater streams change in wet length in response to storm events and seasonal moisture conditions. These low-order channels with temporary flow are pervasive across arid and humid environments yet receive little attention in comparison to perennial waterways. This dissertation examines headwater stream length dynamics at multiple spatial and temporal scales across the Appalachians. I mapped wet stream length in four Appalachian physiographic provinces--the Appalachian Plateau, Blue Ridge, New England, and Valley and Ridge--to characterize seasonal expansion and contraction of the wet network at a broad, regional scale. Conversely, most existing field studies of stream length in headwaters are limited to a single study area or geographic setting. Field mappings showed that wet stream length varies widely within the Appalachians; network dynamics correlated with regional geology as well as local site lithology, geologic structure, and the depth, size, and spatial distribution of surficial sediment deposits. I used the field data to create logistic regression models of the wet network in each physiographic province at high and low runoffs. Topographic metrics derived from elevation data were able to explain the discontinuous pattern of headwater streams at different flow conditions with high classification accuracy. Finally, I used flow intermittency sensors in a single Valley and Ridge catchment to record channel wetting and drying at a high temporal resolution. The sensors indicated stream length hysteresis during storms with low antecedent moisture, with a higher wet network proportion on the rising limb than on the falling limb of events. As a result, maximum network extension can precede peak runoff by minutes to hours. Accurate maps of headwater streams and an understanding of wet network dynamics through time are invaluable for applications surrounding watershed management and environmental policy. These findings will contribute to the burgeoning research on temporary streams and are additionally relevant for studies of runoff generation, biogeochemical cycling, and mass fluxes of material from headwaters.
  • Thumbnail Image
    Characterizing major agricultural land change trends in the Western Corn Belt
    Shao, Yang; Taff, Gregory N.; Ren, Jie; Campbell, James B. Jr. (Elsevier, 2016-12-01)
    In this study we developed annual corn/soybean maps for the Western Corn Belt within the United States using multi-temporal MODIS NDVI products from 2001 to 2015 to support long-term cropland change analysis. Based on the availability of training data (cropland data layer from the USDA-NASS), we designed a cross-validation scheme for 2006–2015 MODIS data to examine the spatial generalization capability of a neural network classifier. Training data points were derived from a three-state subregion consisting of North Dakota, Nebraska, and Iowa. Trained neural networks were applied to the testing sub-region (South Dakota, Kansas, Minnesota, and Missouri) to generate corn/soybean maps. Using a default threshold value (neural network output signalP0.5), the neural networks performed well for South Dakota and Minnesota. Overall accuracy was higher than 80% (kappa > 0.55) for all testing years from 2006 to 2015. However, we observed high variation of classification performance for Kansas (overall accuracy: 0.71–0.82) and Missouri (overall accuracy: 0.65–0.77) for various testing years. We developed a threshold-moving method that decreases/increases threshold values of neural network output signals to match MODIS-derived corn/soybean acreage with the NASS acreage statistics. Over 70% of testing states and years showed improved classification performance compared to the use of a default 0.5 threshold. The largest improvement of kappa value was about 0.08. This threshold-moving method was used to generate MODIS-based annual corn/soybean map products for 2001–2015. A non-parametric Mann-Kendall test was then used to identify areas that showed significant (p < 0.05) upward/downward trends. Areas showing fast increase of corn/soybean intensities were mainly located in North Dakota, South Dakota, and the west portion of Minnesota. The highest annual increase rate for a 5-km moving window was about 6.8%.
  • Thumbnail Image
    Coastal Erosion Hazard in Bangladesh: Space-time pattern analysis and empirical forecasting, impacts on land use/cover, and human risk perception
    Islam, Md Sariful (Virginia Tech, 2023-06-27)
    Coastal areas are vulnerable to different natural hazards, including hurricanes, cyclones, tsunami, floods, coastal erosion, and saltwater intrusion. These hazards cause extensive social, ecological, economic, and human losses. Continued climate change and sea-level rise is expected to substantially impact the people living in coastal areas. Sea level rise poses serious threats for the people living in the coastal zone, which leads to coastal erosion, inundations in the low-lying areas, tidal water encroachment and subsequent salt-water intrusion, as well as the displacement of the people living along the coast. Coastal erosion is one of the biggest environmental threats in the coastal areas globally. In Bangladesh, coastal erosion is a regularly occurring and major destructive process, impacting both human and ecological systems at sea level. The Lower Meghna estuary, located in southern Bangladesh, is among the most vulnerable landscapes in the world to the impacts of coastal erosion. Erosion causes population displacement, loss of productive land area, loss of infrastructure and communication systems, and, most importantly, household livelihoods. For a lower middle-class country, such as Bangladesh, with limited internal resources, it is hard to cope with catastrophic natural hazards, such as coastal erosion and its related consequences. This research aims to advance the scientific understanding of past and future coastal erosion risk and associated changes in land change and land cover using geospatial analysis techniques. It also aims to understand the patterns and drivers of human perception of coastal erosion risk. To place the research questions and objectives in content, Chapter 1 includes a brief introduction and literature review of the coastal erosion context in Bangladesh. Chapter 2 assesses different methods of prediction to investigate the performance of future shoreline position predictions by quantifying how prediction performance varies depending on the time depths of input historical shoreline data and the time horizons of predicted shorelines. Chapter 3 evaluates historical land loss and how well predicted shorelines predict amounts of succeeding LULC resources lost to erosion. Chapter 4 focuses on the patterns and drivers of erosion risk perception using data from spatially explicit measures of coastal erosion risk derived from satellite imagery and a random sample survey of residents living in the coastal communities. In summary, this research advances our scientific understanding of past and future coastal erosion risk and associated changes in land change and land cover using geospatial analysis techniques. It also enhances the understanding of the patterns and drivers of human perception of coastal erosion risk by combining satellite imagery and social survey data. Compared to much of the coastal erosion literature, this work draws from a 35-year time series of satellite-derived shorelines at annual temporal resolution. This time depth enables us to employ a temporal design strategy expected to yield a robust characterization of space-time erosion patterns. This study also enabled us to assess how well predicted shorelines predict amounts of succeeding LULC resources lost to erosion by using long-term historical data. The innovative we use has potential applications to other deltas and vulnerable shorelines globally. While empirical results are specific to the project's study area, results can inform the region's shoreline forecasting ability and associated mitigation and adaptation strategies.
  • Thumbnail Image
    Comparing UAV and Pole Photogrammetry for Monitoring Beach Erosion
    Gonzales, Jack Joseph (Virginia Tech, 2021-09-14)
    Sandy beaches are vulnerable to extreme erosion during large storms, as well as gradual erosion processes over months and years. Without monitoring and adaptation strategies, erosion can put people, homes, and other infrastructure at risk. To effectively manage beach resources and respond to erosion hazards, coastal managers must have a reliable means of surveying the beach to monitor erosion and accretion. These elevation surveys typically incorporate traditional ground-based surveying methods or lidar surveys flown from large, fixed-wing aircraft. While both strategies are effective, advancements in photogrammetric technology offers a new solution for topographic surveying: Structure from Motion (SfM). Using a set of overlapping aerial photographs, the SfM workflow can generate accurate topographic surveys, and promises to provide a fast, inexpensive, and reliable method for routine beach surveying. Unmanned aerial vehicles (UAVs) are often successfully employed for SfM surveys but can be limited by poor weather ad government regulations, which can make flying difficult or impossible. To circumvent these limitations, a digital camera can be attached to a tall pole on a mobile platform to obtain aerial imagery, avoiding the restrictions of UAV flight. This thesis compares these two techniques of image acquisition for routine beach monitoring. Three surveys were conducted at monthly intervals on a beach on the central South Carolina coast, using both UAV and pole photogrammetry. While both methods use the same software and photogrammetric workflow, the UAV produced better results with far fewer processing artifacts compared to pole photogrammetry.
  • Thumbnail Image
    Comparison of EPIC-Simulated and MODIS-Derived Leaf Area Index (LAI) across Multiple Spatial Scales
    Iiames, John S.; Cooter, Ellen; Pilant, Andrew N.; Shao, Yang (MDPI, 2020-08-26)
    Modeled leaf area index (LAI) in conjunction with satellite-derived LAI data streams may be used to support various regional and local scale air quality models for retrospective and future meteorological assessments. The Environmental Policy Integrated Climate (EPIC) model holds promise for providing LAI within a dynamic range for input into climate and air quality models, improving on current LAI distribution assumptions typical within atmospheric modeling. To assess the potential use of EPIC LAI, we first evaluated the Moderate Resolution Imaging Spectroradiometer (MODIS) LAI product collections 5 and 6 (i.e., Mc5, Mc6) with in situ LAI estimates upscaled at four 1.0 km resolution research sites distributed over the Albemarle-Pamlico Basin in North Carolina and Virginia, USA. We then compared the EPIC modeled 12.0 km resolution LAI to aggregated MODIS LAI (Mc5, Mc6) over a 3 × 3 grid (or 36 km × 36 km) centered over the same four research sites. Upscaled in situ LAI comparison with MODIS LAI showed improvement with the newer collection where the Mc5 overestimate of +2.22 LAI was reduced to +0.97 LAI with the Mc6. On three of the four sites, the EPIC/MODIS LAI comparison at 12.0 km resolution grid showed similar weighted mean LAI differences (LAI 1.29–1.34), with both Mc5 and Mc6 exceeding EPIC LAI across most dates. For all four research sites, both MODIS collections showed a positive bias when compared to EPIC LAI, with Mc6 (LAI = 0.40) aligning closer to EPIC than the Mc5 (LAI = 0.61) counterpart. Despite modest differences between both MODIS collections and EPIC LAI, the overestimation trend suggests the potential for EPIC to be used for future meteorological alternative management applications on a regional or national scale.
  • Thumbnail Image
    A Comparison of Imperviousness Derived from a Detailed Land Cover Dataset (DLCD) versus the National Land Cover Dataset (NLCD) at Two Time Periods
    Cooper, Brandon Elliott (Virginia Tech, 2016-09-01)
    To address accuracy concerns of the National Land Cover Dataset (NLCD), this case study compares impervious surface from the NLCD to a Detailed Land Cover Dataset (DLCD) for the Town of Blacksburg, Virginia over two time periods (2005/2006 and 2011) at spatial aggregation scales (fine to coarse) and scopes (site-specific to area-extent). When comparing the total impervious surface area, the NLCD overestimated the DLCD by appreciable amounts (12-27%) for the entire town and across all specified land use zones for both time periods examined. A binary pixel-wise accuracy assessment of impervious surface revealed that the NLCD performed well for all scopes except for the single family land use zone (user accuracy <40%). The spatial aggregation of pixels to 90-m led to improved agreement between the two datasets. Using the DLCD as a reference, an empirical normalization equation was successfully applied to the NLCD to further reduce overestimation and data skewness.
  • Thumbnail Image
    Detecting and Modeling Landfast Ice in the Alaskan Bering Sea
    Jensen, David Aaron (Virginia Tech, 2020-06-19)
    Seasonal sea ice – ice which freezes in late fall and melts completely the following summer – is a central feature in the ecology, geomorphology, and climatology of the Bering Sea. In this region's coastal zones, sea ice becomes locked into a stationary position against the coastlines to become landfast ice, which influences bioegophysical processes in the region, as well as exchanges of energy and matter among land, ocean, and atmosphere. It provides a platform for human mobility and subsistence activities, habitat for certain marine mammals, regulates terregenous nutrient cycling into ocean environments, and modulates the effect of erosive wind/wave action against coastlines. However, a thorough understanding of how this stationary ice, known as landfast ice, affects biogeophysical processes in the Bering Sea is limited by a lack of data on its areal coverage and seasonal duration. This dissertation establishes a baseline set of observations of landfast ice conditions in the Bering Sea through the creation and analysis of continuous spatial datasets. Chapter 1 focuses on the landfast ice annual cycle in the Eastern Bering Sea, which spans from the western tip of the Seward Peninsula to the southernmost point on the Yukon-Kuskokwim River Deltas. Chapter 1 results in the creation of landfast ice spatial data in these areas ranging from 1996 – 2008. Results show the spatial distribution and seasonal duration of landfast ice vary regionally within our study area, does not generally reach water depths associated with stabilization of the landfast ice cover in other regions of the Arctic, and is shortening in seasonal duration by approximately 9 days. Chapter 2 focuses on the landfast ice annual cycle on St. Lawrence Island, an Alaska Island located in the northern Bering Sea. Chapter 2 results in the creation of landfast ice spatial data in these areas ranging from 1996 – 2019. Results show the spatial distribution of landfast ice to vary regionally on the island, based on the coastlines orientation towards prevailing winds that transport pack ice through the Bering Strait. We also observed a sharp decline in landfast ice cover from 2017-2019, which coincides with record-breaking declines in sea ice coverage for the entire Bearing Sea. We also found coastal morphology and orientation have limited explanatory power when modeling landfast ice widths – the distance between the landfast ice edge and coastline – suggesting the consideration of meteorological variables is needed to improve models. Chapter 3 uses the landfast ice data from Chapter 2 to create an explanatory logistic regression model of landfast ice cover on St. Lawrence Island, using a combination of geographic and meteorological predictor variables. Using these variables, the model was able to predict the location of landfast ice cover with 80-90% accuracy, depending on the region of St. Lawrence Island. The model outputs resulted in very low commission error, with high omission error, which may be improved in future studies with the additional predictor variables. Cumulatively, this dissertation is the most comprehensive analysis of landfast ice cover to date on Alaskan Bering Sea coastlines. Research findings advance scholarly understandings of coastal ice conditions in the Bering Sea, and the geographic as wellas meteorological factors that enable their presence.
  • Thumbnail Image
    Detection of Tornado Damage via Convolutional Neural Networks and Unmanned Aerial System Photogrammetry
    Carani, Samuel James (Virginia Tech, 2021-10-21)
    Disaster damage assessments are a critical component to response and recovery operations. In recent years, the field of remote sensing has seen innovations in automated damage assessments and UAS collection capabilities. However, little work has been done to explore the intersection of automated methods and UAS photogrammetry to detect tornado damage. UAS imagery, combined with Structure from Motion (SfM) output, can directly be used to train models to detect tornado damage. In this research, we develop a CNN that can classify tornado damage in forests using SfM-derived orthophotos and digital surface models. The findings indicate that a CNN approach provides a higher accuracy than random forest classification, and that DSM-based derivatives add predictive value over the use of the orthophoto mosaic alone. This method has the potential to fill a gap in tornado damage assessment, as tornadoes that occur in wooded areas are typically difficult to survey on the ground and in the field; an improved record of tornado damage in these areas will improve our understanding of tornado climatology.
  • Thumbnail Image
    Detection of Urban Damage Using Remote Sensing and Machine Learning Algorithms: Revisiting the 2010 Haiti Earthquake
    Cooner, Austin J.; Shao, Yang; Campbell, James B. Jr. (MDPI, 2016-10-20)
    Remote sensing continues to be an invaluable tool in earthquake damage assessments and emergency response. This study evaluates the effectiveness of multilayer feedforward neural networks, radial basis neural networks, and Random Forests in detecting earthquake damage caused by the 2010 Port-au-Prince, Haiti 7.0 moment magnitude (Mw) event. Additionally, textural and structural features including entropy, dissimilarity, Laplacian of Gaussian, and rectangular fit are investigated as key variables for high spatial resolution imagery classification. Our findings show that each of the algorithms achieved nearly a 90% kernel density match using the United Nations Operational Satellite Applications Programme (UNITAR/UNOSAT) dataset as validation. The multilayer feedforward network was able to achieve an error rate below 40% in detecting damaged buildings. Spatial features of texture and structure were far more important in algorithmic classification than spectral information, highlighting the potential for future implementation of machine learning algorithms which use panchromatic or pansharpened imagery alone.
  • Thumbnail Image
    Detection of Urban Damage Using Remote Sensing and Machine Learning Algorithms: Revisiting the 2010 Haiti Earthquake
    Cooner, Austin Jeffrey (Virginia Tech, 2016-12-19)
    Remote sensing continues to be an invaluable tool in earthquake damage assessments and emergency response. This study evaluates the effectiveness of multilayer feedforward neural networks, radial basis neural networks, and Random Forests in detecting earthquake damage caused by the 2010 Port-au-Prince, Haiti 7.0 moment magnitude (Mw) event. Additionally, textural and structural features including entropy, dissimilarity, Laplacian of Gaussian, and rectangular fit are investigated as key variables for high spatial resolution imagery classification. Our findings show that each of the algorithms achieved nearly a 90% kernel density match using the United Nations Operational Satellite Applications Programme (UNITAR/UNOSAT) dataset as validation. The multilayer feedforward network was able to achieve an error rate below 40% in detecting damaged buildings. Spatial features of texture and structure were far more important in algorithmic classification than spectral information, highlighting the potential for future implementation of machine learning algorithms which use panchromatic or pansharpened imagery alone.
  • Thumbnail Image
    Drought and Human Impacts on Land Use and Land Cover Change in a Vietnamese Coastal Area
    Tran, Hoa Thi; Campbell, James B. Jr.; Wynne, Randolph H.; Shao, Yang; Phan, Son Viet (MDPI, 2019-02-08)
    Drought is a dry-weather event characterized by a deficit of water resources in a period of year due to less rainfall than normal or overexploitation. This insidious hazard tends to occur frequently and more intensively in sub-humid regions resulting in changes in the landscape, transitions in agricultural practices and other environmental-social issues. The study area is in the sub-humid region of the northern coastal zone of Binh Thuan province, Vietnam—Tuy Phong district. This area is indicated as a subject of prolonged droughts during 6-month dry seasons, which have occurred more frequently in recent years. Associated with economic transitions in agricultural practicing, urbanization, and industrialization, prolonged droughts have resulted in rapid changes in land use and land cover (LULC) in Tuy Phong, especially in three coastal communes: Binh Thanh, Lien Huong, and Phuoc The. A bi-temporal analysis using high-resolution data, the 2011 WorldView2 and the 2016 GeoEye1, was examined to assess LULC changes from observed severe droughts in those three communes. Results showed a dramatic reduction in the extent of hydrological systems (about 20%), and significant increases of tree canopies in urban areas and near the coastal areas (approximately 76.8%). Paddy fields declined by 51% in 2016; such areas transitioned to inactive status or were alternated for growing drought-tolerant plants. Shrimp farming experienced a recognizable decrease by approximately 44%. The 2014 map and field observations during summer 2016 provide references for object-based classification and validation. Overall agreement of results is about 85%.