ASPRS 2014 Annual Conference & co-located JACIE Workshop

  • Increase font size
  • Default font size
  • Decrease font size


ASPRS 2014 Annual Conference

 Geospatial Power in Our Pockets

& co-located JACIE Workshop
Joint Agency Commercial Imagery Evaluation (JACIE) Workshop

Louisville, Kentucky USA   *  March 23-28, 2014  *  The Galt House Hotel

Join the American Society for Photogrammetry and Remote Sensing(ASPRS) for the 2014 Annual Conference as we head to the home of the Kentucky Derby, the Louisville Slugger baseball bat and Southern Hospitality, Louisville, Kentucky, March 23 - 27, 2014!

This year we are excited to welcome the JACIE Workshop to co-locate in Louisville. The JACIE Workshop will be held March 26 - 28, 2014 at the Galt House Hotel and will be combining a general session and special technical sessions throughout the week with the ASPRS Conference. This is an exciting partnership for both organizations!


The intent of the JACIE workshop is to exchange information regarding the characterization and application of the commercial imagery used by the government. This workshop is focused on the synergy of high, medium and low resolution imagery and remote sensing technologies used by the Government. This workshop is sponsored by the Joint Agency Commercial Imagery Evaluation (JACIE) team, a collaborative group of representatives from the National Aeronautics and Space Administration (NASA), the National Oceanic and Atmospheric Administration (NOAA) the United States Department of Agriculture (USDA) and the United States Geological Survey (USGS).

Tell Me About ASPRS 2014

The conference theme: Power in Our Pockets, refers to the technological power of pocket sized devices in our world today. The conference will focus on the various tools, applications, software and overall abilities of technology in the geospatial industry today.

There are LOTS of changes happening for ASPRS conferences! Here are just a few you will see in 2014:

  •     JACIE Workshop co-location
  •     Unmanned Aerial Systems Showcase
  •     Recruitment Way Table Tops
  •     Increased Exhibitor/Attendee Face-time
  •     New session tracks for practical applications
  •     Redesigned conference programs
  •     Presenter abstracts available online

Who Attends?

More than 1,000 imaging and geospatial information professionals gather from across the nation and from around the globe for ASPRS Annual Conferences. And this year, we are expecting a record attendance with the co-location of the JACIE Workshop.

Attendees are mid- and upper-level imaging and geospatial managers from corporations, government agencies, consultants, educators, reseachers, students and field surveyors.


Louisville, Kentucky
Big City Service with Southern Hospitality!

Nestled on the banks of the Ohio River, Louisville, Kentucky has loads of small-town southern hospitality, a cosmopolitan riverfront district linked to the city’s park system, a diverse arts scene, downtown’s Museum Row on Main, and a nationally recognized foodie mecca.

Louisville, no matter how you pronounce's got something for everyone!

ASPRS and the JACIE Workshop will be holding meetings at the wonderful Galt House Hotel while in Louisville! Click here for more information about hotel accommodations.


T7-Mar 27 9:00

Comparative Analysis Using Multi Sensor Data Integration for Extracting Geotechnical Parameters

Abdulla Al-Rawabdeh, University Of Calgary, Department of Geomatics Engineering

Habib Ayman and He Fangning

Comparative Analysis Using Multi Sensor Data Integration for Extracting Geotechnical Parameters Abdulla Al-Rawabdeh, Ayman Habib, Fangning He University Of Calgary, Department of Geomatics Engineering  Digital Photogrammetry Research group  Geotechnical engineering is a relatively new discipline that has developed rapidly over the past 30 odd years, and deals with a wide spectrum of natural geological materials, ranging from low strength soils to high strength rocks. In many parts of the world, rock mass movements are common and as a result present serious safety and mortality risk to humans; it particularly affects the construction activities. Prevention of large natural landslides is difficult, but common sense and good engineering practice can help to minimize these hazards. The monitoring of dangerous areas is becoming more important due to environmental factors and structural failures. The orientations of surface discontinuities are important indicators for geotechnical processes, mapping, and monitoring of landslides. They provide valuable information about the slope activity and related hazards. Traditionally, the discontinuities orientation measurement (Strike and Dip) was carried out during field surveys using a geological compass and an inclinometer to understand the geological structure and establish a geological history of a region. When unstable rock mass conditions are encountered and no opportunity exists to enter the area of interest, direct collection of data becomes quite difficult. Aside from being inaccessible, this process of manual data collection is also very time consuming and presents as subjective data. Traditional methods of geological data acquisition are prone to errors due to sampling difficulties, being very cumbersome, fragment instrumental errors, and the occurrence of human bias. Nowadays, 3D modeling of objects can be achieved through either passive or active remote sensing systems. Active sensors, such as terrestrial laser scanning systems (TLS) have been extensively used for the acquisition of high accuracy three-dimensional spatial or point cloud data. With the increasing quality, availability, and affordability of point cloud data, there is a wide potential for use in various disciplines. However, the TLS in some cases has limitations during the data collection. This work will focus on integrated (Multi-Sensory) data using TLS and close range photogrammetric data to derive the strike and dip measurements of discontinuity planes. The main target in this work is the development of automatic image processing techniques (Semi-Global Dense Matching) to minimize the field work and hence minimizing time, cost, and eliminating safety hazards. The research will illustrate how using Multi-Sensory data can efficiently generate a point cloud using close range photogrammetry recorded at rock mass located along the road cut west of Calgary, Alberta. The point cloud data generated using Semi-Global Dense Matching technique would be also useful for filling the gaps in the TLS data resulting from occlusions, orientation bias, and truncation. The integrated multi-sensory data using for deriving the strike and dip measurements of discontinuity planes using different data processing techniques, and the results provides a comparative analysis between conventional and proposed methods. The validity of the derived measurements is validated using field measurements.

Development of a remote-sensing based methodology for the identification and classification of Kentucky wetlands

Kelly Watson, Eastern Kentucky University

Daniel Mullins and Nicholas Middleton

An estimated 300,000 acres of wetland exist in Kentucky. However, this number is based primarily on the National Wetlands Inventory, which was carried out in the 1970s. Since this time, the use of satellite-based remote sensing for the identification and classification of wetlands has become an important tool for updating wetland maps. Our research contributes to ongoing efforts aimed at improving geospatial techniques for the identification and examination of wetland characteristics. ASTER, Landsat 7, and Landsat 8 imagery were evaluated in combination with field-verification of wetlands in the Upper Cumberland and Green watersheds of Kentucky. Further analysis was performed using a GIS to eliminate false positives, such as farm ponds. The results of a supervised classification of Kentucky wetlands demonstrates remote sensing is a useful tool for classifying, mapping, and detecting change within wetlands—all of which can be used to both improve wetlands management and aid in conservation efforts. 

Monitoring and Predicting Land Use and Land Cover Change on Cedar River Watershed of Iowa for the Next 30 Years

Xin Hong, University of Northern Iowa

Bingqing Liang

Mapping the pattern of land use and land cover (LULC) and predicting its future trend are important for creating and maintaining a sustainable environment. The current project tends to predict LULC change on Cedar River watershed of Iowa for the next 30 years by analyzing past LULC patterns at three time frames:1990, 2000, and 2010, using satellite images collected by SPOT satellite, ASTER sensor, and Landsat satellite, respectively. After initial processing, these images will be classified using object-oriented algorithms to derive ten classes: water, wetland, forest land, grassland, corn/soybeans, barren land, industrial/commercial, roads, residential, and shadow/no data. The resultant classified images will then be analyzed to predict future LULC patterns using the Idrisi Land Change Modeler built on the Markov-chain algorithm. It is expected the results from this project will help to understand the trend of LULC pattern in heavily farmed watershed of Iowa.

Suitability of Remote Sensing Data for African Dust Mapping in the Florida Keys, U.S.A.

Bandana Kar, University of Southern Mississippi

Grant L. Harley

Atmospheric dust can have substantial negative impacts on social, financial and physical environments. Over the past 40 years, African dust outbreaks have increased in the U.S. due to increased desertification and global climate change. To understand the spatial and temporal variability of African dust outbreaks, high-resolution proxy data of atmospheric dust extending beyond the observational period (past ca. 40 years) are required. An alternative is to use remotely sensed (RS) data.  Given the varying spatial, temporal and spectral resolutions, it is essential to determine suitable RS data for such a study. Recently, the annual growth rings of trees growing in the Florida Keys, U.S.A. were correlated with annual solar radiation and atmospheric dust concentration. The purpose of this study is to determine an appropriate remote sensing data set that would accurately represent dust concentration as represented by tree growth rings. To accomplish this goal, derived products from TOMS AI, Meteosat-IR, SeaWiFS, MODIS, MISR, OMI, AVHRR and VIIRS satellite sensors for dust and aerosol will be fused with dendrochronology data (tree-ring growth data as high-resolution proxy and reference data). By comparing the dust concentration information from each of the RS data with the tree-ring data for different time periods, accuracy and variance threshold of each data set will be determined to identify the best RS data. The study will be conducted for the Florida Keys, one of the areas most heavily impacted by African dust as a first step towards developing a reconstruction of African dust events using tree rings.

Land Cover Classification and Analysis using Radar and Landsat Data in North Central Ethiopia

Haile Tadesse, George Mason University

Agriculture is the main economic sector in Ethiopia. It is highly dependent on rainfall, which is erratic and unpredictable. Land use change and deforestation was a common practice in Ethiopia for many years and it is still one of the current environmental problems. According to different studies, deforestation in Ethiopia is caused due to agricultural expansion, firewood collection and logging, settlement and urbanization. One main concern in the land use planning issues in Ethiopia is the lack of accurate data on land use/cover and deforestation. In different articles there is variability of research results and it is very hard to have a sound land cover or forest cover map. The available research results on deforestation varies depending on the type of techniques used to study the research question. If there is no accurate data on the rate of deforestation and land use issues, it will be hard to have an appropriate plan to tackle deforestation and environmental degradation. Land cover and land use change can be studied by remote sensing techniques. For many years, Landsat and other optical sensors has been used to monitor land cover and deforestation in different part of the world. Optical remote sensing was widely used instrument to classify land-use and land-cover. In tropics and my study area, image classification using optical remote sensing is very hard due to lack of good landsat data. The reason behind this is due to the presence of cloud cover. Most of the images are with poor quality and such poor quality data will produce poor land cover classification results. This problem can be improved by using radar data or active sensors which are available from different sources. These active sensors are not affected by cloud cover and can be relaible source of data for tropics. Recent radar data such as RADARSAT-2 and PALSAR have more than one pollarization and can be used to classify land use/cover. In this research study, radar and Landsat data are used to classify land cover and analyze the accuracy results of these classified maps. Besides this, texture and de-spekling techniques used to see the improvments to the classification accuracy since the orginal radar data have poor land cover classification accuracy results due to speckling issues. This study examines land cover classification and analysis using radar and Landsat data in North Central Ethiopia. Landsat TM from USGS and Palsar radar data from Alaska Satellite facility are used for this study. ERDAS Imagine used for geo-referencing, enhancement, classification and accuracy assessments. Maximum likelihood classifier used for the classification. The analysis focuses on the impact of radar data enhancement methods on classification accuracy assessments. For the accuracy assessment, 12229 pixels used which are different from the training pixels used for the classification. The original radar data has low overall classification accuracy (66%). Urban land cover has 17% and 85% producer and user accuracy respectively. Forest and Agriculture have 52.2 and 78.4% producer accuracy respectively. User accuracy for forest is 49.7 % and agriculture has 55.4% user accuracy. These results indicate it is very hard to use the classification map derived from the original radar data. This poor result is caused due to the presence of speckle in the original radar data and it is important to use data enhancement techniques. To improve this classification accuracy, Speckle suppression and Textures measures were used. The de-speckling methods used in this study are Median, Lee-Sigma and Gamma-Map. Different window sizes were used to see the impact of kernel size on accuracy assessments. Lee-sigma, Gamma Map and Median de-speckling techniques improved the overall accuracy by about 15, 18 and 20 % respectively. Increasing window size beyond 27 by 27 decreases the overall accuracy result in this study area. The maximum overall accuracy achieved in this study by de-speckling method is 86.4 % when used Median at 27 by 27 window size. Urban producer accuracy improved by 58 % by using Median de-speckling at 27 by 27 kernel size. Median de-speckle with window size greater than 19 by 19 improved urban user accuracy to 100% from 85.87% in the original radar data. Overall, all de-speckling techniques improved the urban user accuracy to more than 90%. Gamma Map and Lee-sigma at 27 by 27 kernel size also improved urban producer accuracy by 53 and 33% respectively. Similarly, Median and Gamma Map and Lee-Sigma at 27 by 27 improved user accuracy of forest by 38, 36 and 33 % respectively. Agriculture user accuracy improved by only 5 % when used Lee-Sigma at 27 by 27 window size. Kernel size greater than 19 decreases user accuracy of agriculture when Median and Gamma-Map are applied.  There are many texture measures but for this study only variance is used. Variance texture measure was used using different window size. In this study, the highest overall accuracy result (88.77%) is achieved using window size 51 by 51. This shows 22% improvement as compared to the original radar data (66%). The impact of texture on accuracy varies from one land cover to another. Urban and forest producer accuracy improved by 58 and 26 % respectively at window size 43 by 43. Increasing the window size increased the producer accuracy of agriculture by 21% as compared to the original radar data. However, increasing window size beyond 43 by 43 increases the confusion between urban, forest and agriculture. Urban user accuracy improved to 100 % after using 15 by 15 window size. In general, urban areas can be classified more accurately using variance texture due the diversity of digital numbers (DN). User accuracy improved by 33 and 27% respectively for forest and agriculture when 43 by 43 window size is used.  Landsat data produced an overall classification accuracy of 93.7%. Urban has 98 and 79 % producer and user accuracy results respectively. Even if this result is very good, Landsat data availability is very limited in the study area due to cloud cover. Combining Landsat and derived radar data measures improved accuracy assessment by 5% as compared to the Landsat data. The original radar data produced low overall accuracy and individual land cover accuracy results. This study also shows the importance of texture and de-speckling techniques to improve land cover classification accuracy in radar data. Therefore, radar data can be used as an alternative to Landsat data in tropics and Ethiopia for land cover/use classification.

Geo-Spatial Technologies for Nigerian Urban Security and Crime Management - A Study of Abuja Crime Hotspot Mapping and Analysis

Dr. Matthiew Adepoju, National Space Research and Development Agency

Seidu Mohammed, Halilu Shaba, Mohammed Ozigis, Idris Ibrahim, Blessing Alau, and Seun Adeluyi

Abuja is one of the fastest growing cities in the sub-Sahara Africa. The city lacks the modern management techniques for an effective crime mapping, monitoring and management to meet the attainment of liveable environment despite its aesthetically pleasing outlook of a modern city. The cities of the developed world are managed with the intelligence provided by Geo-spatial technologies. The advances in space technologies have made the onerous task of managing crime possible. However, the availability of these technologies has not been exploited, utilised nor domesticated by various institutions charged with the responsibility of city planning and management to attain secure and safe environment in Nigerian urban and rural communities. Geographic information systems, remote sensing and allied technologies have manifested in various forms in the last four decades particularly since the launch of LANDSAT earth observatory satellites these has provided baseline information for intelligence gathering. The very high resolution images provided by the new generation of satellites have made the integration of GIS/RS for urban crime mapping not only possible but also effective for day-to-day running and management of many aspects of city life. The NigeriaSat-2 and Ortho-rectified Quick-Bird images, basemap, master plan and questionnaires were used to generate the crime dataset which was later aggregated to show the crime hotspots and Coldspots areas within the residential districts of Abuja Federal Capital City, Phase 1. A proximity analysis was later carried out to ascertain the relationship between crime hotspots and Coldspots and police divisional stations, slum settlement as well the various parks and gardens in the study area. The result showed significant correlation between parks and gardens and crime as well as positive correlation between slum settlement and crime in the study area.

Analysis of epipolar geometry mapping relation for digital frame camera images

Zexun Geng, Zhengzhou Insitute of Surveying and Mapping

Qing Xu, Baoming Zhang, Zhihui Gong, and Dazhao Fan

Digital aerial frame cameras are widely used in mapping applications. As one most effective and frequently used constraints in digital photogrammetry research and practices, however, epipolar geometry relation of digital frame camera is not investigated deeply. In this paper, analytical transforming relation between center projective image(CPI) obtained by digital frame camera and epipolar image(EPI) is derived after relative orientation finished. The contributions of this paper are as follows. First the monotone mapping relation from CPI to EPI is obtained, which lays the foundation of two calibration algorithms in epipolar image generation. One is the fast indirect rectification per block algorithm; the other is direct rectification algorithm for any sub-region in CPI. Then a critical epipolar line is proved to be existence in CPI, which makes epipolar image formation process easer understanding. The coordinates of epipolar point that all epipolar lines intersect at in CPI, are deduced thirdly. From that epipolar point, it is very easy to get slant epipolar line that goes through any image point in CPI.

Use of DubaiSat-1 Imagery and In-Situ Observations for Nutrient Monitoring in a Dubai Coastal Area

Tarig Ali, American University of Sharjah

Maruf Mortula and Serter Atabay

The productivity of coastal marine systems is highly affected by nutrient availability. Nutrient enrichment in Dubai coast has been one of the primary environmental concerns because it can lead to eutrophication. The increased algal growth and the extended productivity from eutrophication normally result in increased oxygen consumption by bacteria to decompose the dead algal cells fallen to the bottom leading to low-oxygen water. This can lead to the killing of fish and lowering the biotic diversity. Furthermore, nutrient enrichment is one factor that is linked to the growth of the harmful algal blooms (HAB’s), which has devastated the coasts of UAE and the whole Arabian Gulf region in 2008. The problem presented above calls for a large-scale monitoring program of nutrients since the current programs only provide point-based information, which is insufficient to study the spatial patterns of nutrients and their transport mechanisms. DubaiSat-1 imagery and in-situ observations have been utilized in this study to monitor the nutrients and understands their transport processes at a micro-scale level in a Dubai coastal area.

Advanced Image Processing using Image I/O-Ext  and Java Advanced Imaging(JAI)

Rakesh Kumar Mishra, Department of Geodesy and Geomatics Engineering, University of New Brunswick

Yun Zhang

Image I/O-Ext and Java Advanced Imaging API (JAI) are open source tools that provide a set of object-oriented interfaces for image input/output (I/O) and image processing operations. Image I/O-Ext extends the capabilities of Java Image I/O by leveraging GDAL which is a raster Geospatial Data Abstraction Library capable of managing a large set of raster formats. Therefore, because of GDAL, Image I/O-Ext is able to support many raster formats including BigTiff, ERDAS, ENVI, and ArcGIS. On the other hand, Java Advanced Imaging API (JAI) provides a set of object-oriented interfaces for image processing operations. JAI supports a simple, high-level programming model which allows images to be manipulated easily in Java applications. JAI provides a high-performance, platform-independent, extensible image processing framework. These imaging technologies greatly improves developers' ability to implement portable image processing applications and their flexible, scalable design meets the demands of the geospatial, medical, commercial, network and government imaging markets. Furthermore, JAI has several advantages including cross-platform, highly extensible, powerful, high performance, and interoperable. This paper presents a frame work to use Image I/O-Ext and JAI together for Remote Sensing (RS) image processing programming. Research work in RS image processing area often requires programming in order to realize algorithms developed. The proposed frame work considerably reduces the programming effort and time of a researcher because Image I/O-Ext support seamless reading and writing of raster formats and JAI has many inbuilt functions required for the image processing. This frame work facilitates a researcher to decompose complex image processing algorithms into small, manageable modular components and hence ensures the reusability and scalability of the developed programs. Furthermore, the developed frame work simplifies the creation of imaging applications that support a broad range of systems, from thin clients to powerful workstations.

ZY-3 satellite DEM verification and refinement with SRTM

Yongjun Zhang, Wuhan University

Bo Wang, Yansong Duan, and Qi Chen

As the first high-resolution civil optical satellite, ZY-3 satellite is able to obtain high-resolution multi-view images with three linear array sensors. The images can be used to generate Digital Elevation Models (DEM) through dense matching of stereo images. However, due to the clouds, forest, water and buildings covered on the images, there are some problems in the dense matching results such as outliers and areas failed to be matched (matching holes). This paper introduced an algorithm to verify the accuracy of DEM that generated by ZY-3 satellite with Shuttle Radar Topography Mission (SRTM). Since the accuracy of SRTM (Internal accuracy: 5 m; External accuracy: 15 m) is relatively uniform in the worldwide, it may be used to improve the accuracy of ZY-3 DEM. Based on the analysis of mass DEM and SRTM data, the processing can be divided into two aspects. The registration of ZY-3 DEM and SRTM can be firstly performed using the conjugate line features and area features matched between these two datasets. Then the ZY-3 DEM can be refined by eliminating the matching outliers and filling the matching holes. The matching outliers can be eliminated based on the statistics on Local Vector Binning (LVB). The matching holes can be filled by the elevation interpolated from SRTM. Some works are also conducted for the accuracy statistics of the ZY-3 DEM. The paper includes four aspects: (1) The registration of ZY-3 DEM and SRTM, (2) The elimination of matching outliers, (3) The processing of matching holes, (4) The accuracy statistics of ZY-3 DEM.

Land Cover Mapping at Level 2 or Higher for Multi-county Regional Areas:  Comparison of SPOT 5, SPOT 6, and PLEIADES imagery

Hyo Jin Ahn, Hunter College-CUNY

Sean Ahearn

Land cover mapping at Level 2 or finer for large areas that may include multiple counties and contain both urban and non-urban features is an extremely challenging task.  Aerial photography provides high spatial resolution, but its inconsistent radiometric quality due to the absence of any calibration, results in poor classification accuracy.  This problem is especially pronounced over very large areas ( i.e. > 1,000 sq. mi.) where the development of a rule-base does not apply beyond individual strips or blocks of photography.  Here we examine the use of 2.5 meter SPOT 5 4 band high resolution imagery in combination with derivatives from the Light Detection and Ranging (LiDAR) datasets and other Geographic Information System (GIS) vector datasets for classification of region of over  4,000 square miles in Kansas and Missouri.  An Object Based Image Analysis (OBIA) classification method for classification into 13 land cover classes was carried out with stepwise processing procedure.  We then compare the resulting land cover maps produced by 2.5 m SPOT 5, 1.5 m SPOT 6 and 0.5 m PLEIADES data for a 25 square kilometer test site.  We examine the affect of the differing spectral bands and their derived ratios; and the affect of spatial resolution on classification accuracy.

T7-Mar 27 9:00

Airborne LiDAR for evaluating the impacts of urbanization on forest carbon dynamics: A case study in the City of Charlotte

Christopher Godwin, Student of The University of North Carolina at Charlotte

Gang Chen and Singh Kunwar

Healthy urban forests play a key role in building human resilience to climate change. They serve as a natural carbon sink that reduces CO2 emissions while enhancing other ecosystem services to create more livable cities. Presently, urban forests are affected by different urbanization patterns that vary across cities and local communities. Compared to the city-level analysis, quantification of fine-scale community-level carbon dynamics facilitates precise forest management along urbanizing landscapes. In this study, we assess the impacts of urbanization on forest carbon storage and sequestration over time at the community level, using Charlotte as a case study site. Specifically, we analyzed: 1) How different residential development patterns impact the rate of carbon sequestration as trees grow over time? And 2) how much can forest growth (accumulation of carbon) compensate for the carbon loss due to rapid urban expansion?  To address these questions, multitemporal (2007 and 2012) small footprint airborne LiDAR data were used to retrieve forest above-ground biomass in a wide range of Charlotte communities, representing multiple development patterns (e.g., low, medium, and high residential density). We evaluated the difference in carbon loss during the five-year (2007-2012) urban expansion. We also calculated the rate of carbon sequestration among the communities developed in different periods from the mid-nineteenth century to the recent. Preliminary results indicate that community patterns have a major impact on forest carbon storage and the rate of carbon sequestration. Although urban expansion has caused substantial carbon losses, certain development patterns could facilitate trees to mature and reach a higher carbon storage level than the others. Hence, sustainable community development should accommodate both the space needs from human and urban trees.  

Improving techniques for historic urban tree cover mapping with archival moderate resolution remote sensing data

Andrew Johnston, Smithsonian Institution

Urban forests are a central element of the urban environment. Improved observations of historic tree cover dynamics are required to better understand how the urban environment changes though time. In this study, satellite remote sensing techniques were applied to observe past variability in tree cover area within the District of Columbia using highly calibrated Landsat data. Validation was performed with data from field surveys and public geospatial data on standing tree cover. Testing of alternate methodologies demonstrated that an approach utilizing support vector regression produced results with greater accuracy across the city when compared to linear spectral mixture analysis. Per-pixel uncertainty remained high using both techniques. Spectral mixture analysis overestimated tree cover in low population density areas and underestimated tree cover in the urban core, while support vector regression provided consistent accuracy across land use types. The consistent reliability allowed results from support vector regression to be used for observing tree cover changes between different land use zones. This made it possible to identify past tree cover changes in low density residential zones within the District of Columbia. These results provide useful background information for maintenance and resource management as part of efforts to monitor and expand urban tree cover. Further development of these methods may enable their application with archival moderate resolution satellite remote sensing data for other study areas.

Downscaling On Demand:  Using Web Services to Simulate High-Resolution Canopy Structure

Gordon Green, City University of New York

Sean C. Ahearn and Wenge Ni-Meister

While increasing storage and processing capacity have made it possible for applications based on remotely sensed data to operate at higher and higher spatial and temporal resolutions, the increasing connectivity between datasets has similarly expanded the available options.  Web mapping services make remote sensing data available on demand such that datasets need not be downloaded and processed in their entirety, or in isolated static subsets.  Rather, algorithms can be applied as needed on the fly, by combining only the exact data required to process any given request.  This approach is particularly useful for applications where the desired spatial or temporal resolution is too high to gather or process the data in wall-to-wall form, and where localized regions of interest can be usefully estimated as needed.  <br /> <br /> Local biomass estimation is an example of an application that could benefit from local high-resolution estimates of canopy structure in areas where direct measurements are not available.  To support this and other such applications, we present a downscaling methodology that simulates 1 m-resolution canopy structure at an individual-tree level from globally-available remote sensing datasets, using an on-demand web-service-based architecture, building on methods implemented for the coterminous United States as described in Green et al., 2013 (PE&RS). Data from Landsat, MODIS, and ICESat-based data products are fused with high-resolution airborne lidar data to simulate high-resolution canopy models (calibrated with airborne lidar), which can act as an estimate of canopy structure in the absence of actual lidar or other height data, and which can be used to perform volume-based downscaling of existing lower-resolution biomass datasets. 

You are here: Home Louisville 2014