A Spatial-Temporal Dynamic Research Questions: … · 2016-08-23 · 4B 5B 6B. 7B 8B 9B 10B. Year....
1
A Spatial-Temporal Representation of Land Subsidence in the Northwest Phoenix Valley, Arizona Brian R. Sovik, GISP School of Geographical Sciences and Urban Planning Arizona State University Study Area A Study Area B Background Results Conclusion There are currently 26 identified active land subsidence features in the state of Arizona covering over 2,000 square miles. 1 By means of spatial query, 15 of the 26 land subsidence features are partially or completely within a few of Arizona’s Urbanized Areas (Avondale, Phoenix - - Mesa, and Tucson have known land subsidence features). 2 This is important because the impact from land subsidence has serious potential consequences: “The results of land subsidence are increased susceptibility to flooding, structural damages (buildings, roads and highways, railroads, flood control structures, well casings, gas and water pipes, transmission lines, electric and gas substations, and sewer lines), flow reversal in drains, sewers, canals, irrigation systems, and aquifers and ground fractures (desiccation cracks, giant desiccation cracks and earth fissures”. 3 FIGURE 1 MAP OF KNOWN SUBSIDENCE FEATURES AND RESEARCH STUDY AREAS Dynamic Phenomena: Research Questions: Space and Time GIS: 0 1 2 3 4 5 6 7 8 1A 2A 3A 4A 5A 6A 7A 8A 9A 10A 11A 12A 13A 14A 1B 2B 3B 4B 5B 6B 7B 8B 9B 10B Downward Subsidence Velocity in Centimeters Per Year Station Location for Cross Sections A and B (1 Mile Spacing) Suface Velocity Comparison Over Time and Location Long Medium Short Land Survey TerraSAR-X Interferometric Synthetic Aperture Radar: Study Areas: Methods: Data Processing: FIGURE 6 FOR CROSS STUDY AREA B THE FOLLOWING WERE DERIVED FROM INSAR PROCESSING: (A). SINGLE LOOK COMPLEX SCENE (2010-08- 20), (B.) COHERENCE IMAGE (2010-08-20 AND 2010-01-23) (C.) COLOR INTERFEROGRAM (D.) DIFFERENTIAL-INTERFEROGRAM (E.) DIFFERENTIAL- INTERFEROGRAM WITH CROSS-SECTION “B” OVERLAY (F.) VERTICAL DISPLACEMENT MAP WITH CROSS-SECTION “B” OVERLAY Source Date Range Number of Days Study Area Temporal Coverage 2004-03-08 to 2010-02-15 2,171 A Long 2008-02-11 to 2010-02-15 736 A Medium 2009-03-02 to 2010-02-15 378 A Short 1957-1997 12,000 A Extremely Long 2004-02-05 to 2010-02-18 2,206 B Long 2008-02-14 to 2010-02-18 736 B Medium 2009-01-29 to 2010-02-18 386 B Short 2010-03-08 to 08-20-2010 165 B Extremely Short Background: ESRI Imagery ArcGlobe Oblique View of Northwest Phoenix, Arizona A frontier in GIScience is converting objective data of dynamic and complex real world phenomena into spatial-temporal building blocks for generating new knowledge. My study includes the analysis of objective data for a complex and dynamic phenomena called land subsidence. The analysis of land subsidence and its temporal state are considered within two known subsidence features within Arizona. Questions my research will attempt to resolve include: 1. Can the temporal state of land subsidence be derived from InSAR data processing? 2. What are the query limitations of GIS both spatially and temporally for the derived results? 3. Can this information then be effectively represented with GIS to in the hope to simplified subjective analysis to advance our understanding of land subsidence? FIGURE 7 FOR STUDY AREA A SHOWING RANK ORDER CHANGES IN VELOCITY POLYGONS. FIGURE 8 THE GRAPH ABOVE SHOWS LAND SUBSIDENCE VELOCITY FOR ALL SAMPLE POINTS FOR CROSS-SECTIONS A AND B. EACH LINE COLOR REPRESENTS DIFFERENT SAMPLE TIME DURATIONS. Results: Each study area received digitized land subsidence polygons, cross-sections, and sample location points separated by 1 mile. The temporal durations between InSAR processing ranged between 165 days to over 12,000 days. Space-time GIS is not an easy task as both are continuous and too rich in detail to fully capture without generalizations. Donna Peuquet has offered the “Space-time Triad” as part of a theoretical framework that three queries GIS should be solvable when dealing with spatial- temporal data in an advanced system: 1. When + Where → What 2. When + What → Where 3. Where + What → When “Topographic mapping with InSAR relies on acquiring data from two different look angles and assumes that the scene imaged did not move between data acquisitions. If the look angles of multiple data acquisitions are identical, equivalent to having a zero baseline, there is no sensitivity to topography, and the interferogram can be used to extract information regarding scene dynamics” 5 Study Area “A” is located to the Northwest of Phoenix and includes Sun City, Sun City West, Surprise, portions of Peoria, Glendale and Luke Air Force Base. The location for this study was selected as it encompasses the “Western Metropolitan Phoenix Land Subsidence feature”. Study Area “B” is located within the McMullen Valley groundwater basin and is roughly 60 miles to the northwest of Study Area A. The main community within the study area is Wendon, Arizona. The listed source documents were required for each study area. An additional requirement was that InSAR was used to compute land subsidence. 7 For consistency, both sites were to be processed the same. A controlled variation was added to each Study Area with an additional data source. Study Area “A” included field survey land subsidence information from 1957-1991. Study Area “B” incorporated the processing of Level0 (raw amplitude and phase data) from TerraSAR-X for the creation of InSAR. At the core, a total of six maps from ADWR showing land subsidence were used as a secondary source. FIGURE 3 ERDAS IMAGINE 2010 RADAR SUITE – INFORMATION 6 The electromagnetic wavelengths used in this research was 3 cm. Scene coverage was close to identical. The figure shown to the left depicts the geometric organization of the satellite, orbital position, and ground solution. The difference resulting from the two image captures minus terrain effects provides a highly accurate surface change solution. FIGURE 2 WHAT, WHEN, WHERE (PEUQUET, 2002) 4 Also important, pattern recognition is paramount for the analysis of spatial-temporal knowledge creation. The following patterns will be applied for investigating land subsidence velocity: Spatial: REGULAR, CLUSTERED, CHAOTIC, RANDOM Temporal: STEADY STATE, OCILLATING, CHAOTIC, RANDOM Study Area A Study Area B Land Subsidence in Western Metropolitan Phoenix Based on ADWR EnviSat Time-Series InSAR Data Time Period of Analysis: 5.9 Years 03/08/2004 To 02/15/2010 Land Subsidence in the McMullen Valley, Maricopa and La Paz Counties Based on ADWR EnviSat Time- Series InSAR Data Time Period of Analysis: 6.0 Years 02/05/2004 To 02/18/2010 Land Subsidence in Western Metropolitan Phoenix Based on ADWR EnviSat Time-Series InSAR Data Time Period of Analysis: 2.0 Years 02/11/2008 To 02/15/2010 Land Subsidence in the McMullen Valley, Maricopa and La Paz Counties Based on ADWR EnviSat Time- Series InSAR Data Time Period of Analysis: 2.0 Years 02/14/2004 To 02/18/2010 Land Subsidence in Western Metropolitan Phoenix Based on ADWR EnviSat Time-Series InSAR Data Time Period of Analysis: 0.9 Years 03/02/2004 To 02/15/2010 Land Subsidence in the McMullen Valley, Maricopa and La Paz Counties Based on ADWR EnviSat Time- Series InSAR Data Time Period of Analysis: 1.0 Years 01/29/2009 To 02/18/2010 OFR Report 94-532 Figure 2. Land subsidence in part of the western Salt River Valley, 1957-1991. 7 TerraSAR-X Amplitude and Phase - Single Look Complex Scenes. 2010-01-23, 2010-03-08, 2010-08- 20. Primary and secondary observational data were processed in this experiment. Cross-section were used to generate one mile stationing. This was to done to enable a profile of information to be generated. Digitized polygons were provided attribution based on their source document. Velocity was calculated in centimeters to maintain conformity with source documents. TABLE 1 SOURCE DOCUMENTS Both study areas had three temporally interdependent sets of land subsidence velocity data representing their respective land subsidence features. Temporally, the various time samples should match if land subsidence velocity is in a steady state for any given location along the tested cross- sections. The behavior of land subsidence was determined to be spatially clustered. The multiple time spans showed how trending in velocity at some locations changed abruptly and were not consistent for each location. These micro- changes in land subsidence velocity if missed could be problematic for at risk critical infrastructure as the results were temporally chaotic. Space-time TRIAD 4 questions were mixed as software limitations provide query capabilities to answer When + Where → What. The representation of land subsidence, although not new to GIS has now been reviewed over multiple interdependent timeframes with a focus on its velocity. Figures 7 and 8 effectively represent simplified subjective analysis to advance our understanding of land subsidence. FIGURE 4 RESEARCH STUDY AREAS TABLE 2 TEMPORAL RANGES FOR DATA USED IN RESEARCH FIGURE 5 CROSS SECTIONS FOR STUDY AREAS AND VELOCITY OF LAND SUBSIDENCE References: 1, 2,7 Arizona Deparment of Water Resources, http://www.azwater.gov/azdwr/GIS/ (last accessed December 5, 2010: 6:42 pm) 3 Budhu, Muniram, Adiyaman, Ibrahim Bahadir. 2009. Mechanics of land subsidence due to groundwater pumping International Journal for Numerical and Analytical Methods in Geomechanics 34:1459-1478. 4 Peuquet, Donna J. 2002. Representations of Space and Time. The Guilford Press. 5 Madsen, S., Zebker, H. 1998. Imaging Radar Interferometry. Principles & Applications of Imaging Radar: manual of Remote Sensing Third Edition, Vol. 2. ed. Henderson, Floyd, Lewis, Anthony. John Wiley and Sons. 6 ERDAS 2010, 2009. Radar Interferometry User Guide. Norcross, GA. ERDAS, Inc. 7 Schumann, Herbert H. Land Subsidence and Earth-Fissure hazards Near Luke Air Force Base, Arizona. 1995. U.S. Geological Survey Subsidence Interest Group Conference, Edwards Air Force Base, Antelope valley, California, November 18-19, 1992: Abstracts and Summary. ed. Prince, Keith R., Galloway, Devin L., Leake, Stanley A., pp. 18-21. InSAR Processing: GIS Representation: Velocity Graph:
Transcript of A Spatial-Temporal Dynamic Research Questions: … · 2016-08-23 · 4B 5B 6B. 7B 8B 9B 10B. Year....
A Spatial-Temporal Representation of Land
Subsidence in the Northwest Phoenix Valley, Arizona
Brian R. Sovik, GISP
School of Geographical Sciences and Urban PlanningArizona State University
Study Area AStudy Area B
Background Results
Conclusion
There are currently 26 identified active landsubsidence features in the state of Arizonacovering over 2,000 square miles.1 Bymeans of spatial query, 15 of the 26 landsubsidence features are partially orcompletely within a few of Arizona’sUrbanized Areas (Avondale, Phoenix - -Mesa, and Tucson have known landsubsidence features).2 This is importantbecause the impact from land subsidencehas serious potential consequences:
“The results of land subsidence are increased susceptibility to flooding, structural damages(buildings, roads and highways, railroads, flood control structures, well casings, gas and waterpipes, transmission lines, electric and gas substations, and sewer lines), flow reversal in drains,sewers, canals, irrigation systems, and aquifers and ground fractures (desiccation cracks, giantdesiccation cracks and earth fissures”.3
FIGURE 1 MAP OF KNOWN SUBSIDENCE FEATURES AND RESEARCH STUDY AREAS
Dynamic Phenomena:
Research Questions:
Space and Time GIS:
0
1
2
3
4
5
6
7
8
1A 2A 3A 4A 5A 6A 7A 8A 9A 10A
11A
12A
13A
14A 1B 2B 3B 4B 5B 6B 7B 8B 9B 10B
Dow
nwar
d Su
bsid
ence
Vel
ocity
in C
entim
eter
s Pe
r Ye
ar
Station Location for Cross Sections A and B (1 Mile Spacing)
Suface Velocity Comparison Over Time and Location
Long
Medium
Short
Land Survey
TerraSAR-X
Interferometric Synthetic Aperture Radar:
Study Areas:Methods:
Data Processing:
FIGURE 6 FOR CROSS STUDY AREA B THE FOLLOWING WERE DERIVEDFROM INSAR PROCESSING: (A). SINGLE LOOK COMPLEX SCENE (2010-08-20), (B.) COHERENCE IMAGE (2010-08-20 AND 2010-01-23) (C.) COLORINTERFEROGRAM (D.) DIFFERENTIAL-INTERFEROGRAM (E.) DIFFERENTIAL-INTERFEROGRAM WITH CROSS-SECTION “B” OVERLAY (F.) VERTICALDISPLACEMENT MAP WITH CROSS-SECTION “B” OVERLAY
Source Date Range Number of Days Study Area Temporal Coverage
2004-03-08 to 2010-02-15 2,171 A Long2008-02-11 to 2010-02-15 736 A Medium2009-03-02 to 2010-02-15 378 A Short1957-1997 12,000 A Extremely Long2004-02-05 to 2010-02-18 2,206 B Long2008-02-14 to 2010-02-18 736 B Medium2009-01-29 to 2010-02-18 386 B Short2010-03-08 to 08-20-2010 165 B Extremely Short
Background: ESRI Imagery ArcGlobe Oblique View of Northwest Phoenix, Arizona
A frontier in GIScience is converting objective data of dynamic and complex real worldphenomena into spatial-temporal building blocks for generating new knowledge. My studyincludes the analysis of objective data for a complex and dynamic phenomena called landsubsidence. The analysis of land subsidence and its temporal state are considered withintwo known subsidence features within Arizona. Questions my research will attempt toresolve include:
1. Can the temporal state of land subsidence be derived fromInSAR data processing?
2. What are the query limitations of GIS both spatially andtemporally for the derived results?
3. Can this information then be effectively represented withGIS to in the hope to simplified subjective analysis toadvance our understanding of land subsidence?
FIGURE 7 FOR STUDY AREA A SHOWING RANK ORDER CHANGES IN VELOCITYPOLYGONS.
FIGURE 8 THE GRAPH ABOVE SHOWS LAND SUBSIDENCE VELOCITY FOR ALLSAMPLE POINTS FOR CROSS-SECTIONS A AND B. EACH LINE COLORREPRESENTS DIFFERENT SAMPLE TIME DURATIONS.
Results:
Each study area received digitized landsubsidence polygons, cross-sections, andsample location points separated by 1 mile.The temporal durations between InSARprocessing ranged between 165 days toover 12,000 days.
Space-time GIS is not an easy task as both are continuous and too rich in detail to fullycapture without generalizations. Donna Peuquet has offered the “Space-time Triad” as part ofa theoretical framework that three queries GIS should be solvable when dealing with spatial-temporal data in an advanced system:
1. When + Where → What2. When + What → Where3. Where + What → When
“Topographic mapping with InSAR relies on acquiring data from two different look angles andassumes that the scene imaged did not move between data acquisitions. If the look anglesof multiple data acquisitions are identical, equivalent to having a zero baseline, there is nosensitivity to topography, and the interferogram can be used to extract information regardingscene dynamics” 5
Study Area “A” is located to theNorthwest of Phoenix and includesSun City, Sun City West, Surprise,portions of Peoria, Glendale and LukeAir Force Base. The location for thisstudy was selected as it encompassesthe “Western Metropolitan PhoenixLand Subsidence feature”.
Study Area “B” is located within theMcMullen Valley groundwater basinand is roughly 60 miles to thenorthwest of Study Area A. The maincommunity within the study area isWendon, Arizona.
The listed source documents were required for each study area. An additional requirementwas that InSAR was used to compute land subsidence. 7 For consistency, both sites were tobe processed the same. A controlled variation was added to each Study Area with anadditional data source. Study Area “A” included field survey land subsidence information from1957-1991. Study Area “B” incorporated the processing of Level0 (raw amplitude and phasedata) from TerraSAR-X for the creation of InSAR. At the core, a total of six maps from ADWRshowing land subsidence were used as a secondary source.
FIGURE 3 ERDAS IMAGINE 2010 RADAR SUITE – INFORMATION6
The electromagnetic wavelengthsused in this research was 3 cm.Scene coverage was close toidentical. The figure shown to theleft depicts the geometricorganization of the satellite, orbitalposition, and ground solution. Thedifference resulting from the twoimage captures minus terrain effectsprovides a highly accurate surfacechange solution.
FIGURE 2 WHAT, WHEN, WHERE (PEUQUET, 2002)4
Also important, pattern recognition is paramount for the analysis of spatial-temporalknowledge creation. The following patterns will be applied for investigating land subsidencevelocity:
Spatial: REGULAR, CLUSTERED, CHAOTIC, RANDOMTemporal: STEADY STATE, OCILLATING, CHAOTIC, RANDOM
Study Area A Study Area B
Land Subsidence in Western Metropolitan Phoenix Based on ADWR EnviSat Time-Series InSAR Data Time Period of Analysis: 5.9 Years 03/08/2004 To 02/15/2010
Land Subsidence in the McMullen Valley, Maricopa and La Paz Counties Based on ADWR EnviSat Time-Series InSAR Data Time Period of Analysis: 6.0 Years 02/05/2004 To 02/18/2010
Land Subsidence in Western Metropolitan Phoenix Based on ADWR EnviSat Time-Series InSAR Data Time Period of Analysis: 2.0 Years 02/11/2008 To 02/15/2010
Land Subsidence in the McMullen Valley, Maricopa and La Paz Counties Based on ADWR EnviSat Time-Series InSAR Data Time Period of Analysis: 2.0 Years 02/14/2004 To 02/18/2010
Land Subsidence in Western Metropolitan Phoenix Based on ADWR EnviSat Time-Series InSAR Data Time Period of Analysis: 0.9 Years 03/02/2004 To 02/15/2010
Land Subsidence in the McMullen Valley, Maricopa and La Paz Counties Based on ADWR EnviSat Time-Series InSAR Data Time Period of Analysis: 1.0 Years 01/29/2009 To 02/18/2010
OFR Report 94-532 Figure 2. Land subsidence in part of the western Salt River Valley, 1957-1991.7
TerraSAR-X Amplitude and Phase - Single Look Complex Scenes. 2010-01-23, 2010-03-08, 2010-08-20.
Primary and secondary observationaldata were processed in thisexperiment. Cross-section were usedto generate one mile stationing. Thiswas to done to enable a profile ofinformation to be generated. Digitizedpolygons were provided attributionbased on their source document.Velocity was calculated in centimetersto maintain conformity with sourcedocuments.
TABLE 1 SOURCE DOCUMENTS
Both study areas had three temporallyinterdependent sets of land subsidencevelocity data representing their respectiveland subsidence features. Temporally, thevarious time samples should match if landsubsidence velocity is in a steady state forany given location along the tested cross-sections.
The behavior of land subsidence was determined to be spatially clustered.The multiple time spans showed how trending in velocity at some locationschanged abruptly and were not consistent for each location. These micro-changes in land subsidence velocity if missed could be problematic for atrisk critical infrastructure as the results were temporally chaotic.
Space-time TRIAD4 questions were mixed as software limitations providequery capabilities to answer When + Where → What.
The representation of land subsidence, although not new to GIS has nowbeen reviewed over multiple interdependent timeframes with a focus on itsvelocity. Figures 7 and 8 effectively represent simplified subjectiveanalysis to advance our understanding of land subsidence.
FIGURE 4 RESEARCH STUDY AREAS
TABLE 2 TEMPORAL RANGES FOR DATA USED IN RESEARCH
FIGURE 5 CROSS SECTIONS FOR STUDY AREAS AND VELOCITY OF LANDSUBSIDENCE
References:1, 2,7 Arizona Deparment of Water Resources, http://www.azwater.gov/azdwr/GIS/ (last accessed December 5, 2010: 6:42 pm)3 Budhu, Muniram, Adiyaman, Ibrahim Bahadir. 2009. Mechanics of land subsidence due to groundwater pumping International Journal for Numerical and Analytical Methods in Geomechanics 34:1459-1478.4 Peuquet, Donna J. 2002. Representations of Space and Time. The Guilford Press.5 Madsen, S., Zebker, H. 1998. Imaging Radar Interferometry. Principles & Applications of Imaging Radar: manual of Remote Sensing Third Edition, Vol. 2. ed. Henderson, Floyd, Lewis, Anthony. John Wiley and Sons. 6 ERDAS 2010, 2009. Radar Interferometry User Guide. Norcross, GA. ERDAS, Inc. 7 Schumann, Herbert H. Land Subsidence and Earth-Fissure hazards Near Luke Air Force Base, Arizona. 1995. U.S. Geological Survey Subsidence Interest Group Conference, Edwards Air Force Base, Antelope valley, California, November 18-19, 1992: Abstracts and Summary. ed. Prince, Keith R., Galloway, Devin L., Leake, Stanley A., pp. 18-21.
InSAR Processing:GIS Representation:
Velocity Graph:
![10B-LR 10B-SUBold.bryston.com/PDF/Manuals/300001[10B].pdf · The 10B-STD and 10B-SUB crossovers generate a summed low pass output signal by first summing or adding together the left](https://static.fdocuments.in/doc/165x107/5fca308acddab466873f1279/10b-lr-10b-10bpdf-the-10b-std-and-10b-sub-crossovers-generate-a-summed-low.jpg)
![10B-LR 10B-SUB - Bryston10B].pdf · The 10B crossover is available in three stock versions; 10B-SUB incorporating frequencies more ... MONO LOW PASS MODE (10B-SUB AND 10B-STD ONLY):](https://static.fdocuments.in/doc/165x107/5afd7a367f8b9a434e8d9dda/10b-lr-10b-sub-10bpdfthe-10b-crossover-is-available-in-three-stock-versions.jpg)