CC BY
94
УДК 528.9:004
Эмиль Р. Байрамов
British Petroleum, GIS Департамент
Рафаэль В. Байрамов
Бакинский государственный университет
Эмин А. Хамидов
Баку, Азербайджан
ИСПОЛЬЗОВАНИЕ ГИС ТЕХНОЛОГИЙ И ДИСТАНЦИОННОГО ЗОНДИРОВАНИЯ ДЛЯ МОНИТОРИНГА ПРОЦЕССОВ ВОССТАНОВЛЕНИЯ БИОЛОГИЧЕСКОГО РАЗНООБРАЗИЯ НАРУШЕННЫХ СТРОИТЕЛЬНОЙ ДЕЯТЕЛЬНОСТЬЮ ЗЕМЕЛЬ
Emil R Bayramov
British Petroleum, GIS Department
Rafael V. Bayramov
Baku State University
Emin A. Hamidov
Baku, Azerbaijan
ENVIRONMENTAL MONITORING OF BIO-RESTORATION PROCESSES FOR DISTURBED LANDS BY HUMAN CONSTRUCTION ACTIVITIES USING GIS AND REMOTE SENSING
Азербайджан - страна нефтяных и газовых ресурсов и это является причиной влияния различных строительных процессов на окружающую среду. Для экспорта нефти и газа используются трубопроводы. В процессе их прокладки было нарушено биологическое разнообразие вдоль трубопроводов. Возникла необходимость восстановления нарушенные земель и возврат их в прежнее состояние, и при этом должны исчезнуть следы строительной деятельности.
Основной целью данного исследования была разработка ГИС и методологий дистанционного зондирования для мониторинга процесса восстановления природного разнообразия, нарушенного в период строительной деятельности. Для проведения этого исследования использовались космические снимки высокого разрешения с применением NDVI-метода, требуемого для мониторинга прироста растений вдоль коридора трубопровода. Результаты анализов показали возможность использования ГИС технологий и дистанционного зондирования для вычисления эффективности процесса восстановления растительного покрова с дальнейшим представлением результатов с помощью ГИС-визуализации и определением участков, где необходимо продолжить работы по восстановлению биологического разнообразия.
Keywords: GIS, Remote Sensing, Transect, NDVI, GPS, IKONOS, FORMOSAT
Introduction
The main goal of this research is to develop GIS based methodology for biorestoration monitoring. Azerbaijan has few recently constructed pipelines which still have a need for bio-restoration of their footprints. To make this work efficient dealing with relatively small ranges of research objects, it is inevitable to apply to GIS and remote sensing technologies. One of the main peculiarities of this research is that it deals with quite a detailed scale of territory and that’s a reason why it was applied to
high resolution and accuracy multi-spectral satellite images (Emil R. Bayramov, 2008). Dealing with this detailed scale of territory it is necessary to consider all possible remote sensing corrections to avoid discrepancies in the final results. The final objective and goal in this research is to develop visual representation of vegetation cover based on vegetation index values and estimated transects and evaluate the trend of vegetation cover change between 2007 and 2008 years.
Materials and methods
Materials applied for this research are IKONOS and FORMOSAT satellite images of 2007 and 2008 years, measured GPS coordinates of transects, ground control points, digital elevation model, habitat, land-use. The following stages of on Fig. 4:
Stagel. Collection of transects along the pipeline. It is based on the count of vegetation cover percentage within a special rectangular quadrant. At the same time, GPS coordinates of transects were measured and converted to GIS. It is necessary to point out that collection of images and transects must be done in the same time frame otherwise separation in time yields additional error. The sample of transect is represented on Fig. 1.
Fig. 1. Transect Measurement
Stage 2. Geometric correction of satellite images was implemented using Ground Control Points and Digital Elevation Models. Quality control of Ortho-photo images using ground control and check points (ERDAS 2000).
Stage 3. Atmospheric Correction for some distorted images. Some of satellite images were distorted during acquisition and appropriate correction were applied.
Stage 4. Calculation of Vegetation Index using IKONOS, FORMOSAT images based on the formula NDVI = (NIR - VIS) / (NIR + VIS) was implemented (Bayramov E. R. 2004) Since the sequence of bands in IKONOS and FORMOSAT is 1 -Red, 2- Green, 3- Blue and 4 -Near Infrared, it means that NDVI = (4 - 1) / (4 + 1).
Stage 5. Calculation of NDVI for FORMOSAT and IKONOS images showed that they have different ranges within 0-1 and it was also identified that Atmospheric effects had an impact to acquired NDVI that is a reason why a special atmospheric correction algorithms were applied to distorted images (ERDAS 2000).
Stage 6. On this stage vegetation cover percentage for transects are correlated with NDVI values extracted from satellite images. This helped to make regression between field vegetation cover percentages and NDVI values acquired from satellite images (Graph 1).
y=0.0013x + 0.034 R:= 0.3466
GENERAL HABITAT 2008
0.400
0 350 -
0.300 -
0 250 -
> 0 200 •
n
z 0150 -
0.100 -
0.050 -
0.000 •
-0.050
*
* l
* » * ♦
* * *
* * t t ♦ * + + 1 , I ~ * * S S * # * *“¥ T * *“ t
l.t * ♦ . _ * ♦ ♦ ♦ *
; ■ ■ ■ ■ —tO 20 30 40— —50 60— —Z0 00 90— K
VC
Graph 1. Regression between NDVI and Vegetation Cover for 2008
Stage 7. Based on the acquired equation, it was possible to recalculate NDVI values to Vegetation Cover percentages. This calculation was applied for recalculation of vector data to Vegetation Cover. The vegetation cover of 2007 is represented on Fig. 2 and of 2008 is represented on Fig. 3.
Fig.2: Vegetation Cover 2007 Fig.3: Vegetation Cover 2008
Stage 8. On this stage using calculated Vegetation Cover values from 2007 and 2008 images, it was applied to calculations to make the statistical graph showing the re-growth between two years. To average NDVI along the entire route it was necessary to divide it into the rational rectangles and extract mean NDVI for each of them. It supported to generalize calculations along the route in a statistical form represented on Graph 2.
4.4 « ,n 3J ^
0.4 0.6 1-2 ffi r .—■ I—II—I i . I l I III fi S\ IT I
__________________________Trend Range {%')_________
Graph 2: Trend between 2007 and 2008
Fig. 4: Workflow for GIS based environmental monitoring
Results
Images with improved accuracy allowed to overlay transects correctly what was inevitable when research deals with similar high scale. Atmospheric correction of some images supported in improvement of accurateness of NDVI acquired from satellite images. Calibrated NDVI for IKONOS and FORMOSAT images gave possibility to adjust them to the same range. Calibrated NDVI was correlated with vegetation cover from transects. Further on calibrated NDVI gave possibility to recalculate normalized vegetation index to vegetation cover values which is more
understandable from environmental analysis aspects. Classification of NDVI values into vegetation cover range gave clear representation of areas with vegetation cover sufficiency and rareness. As a result, vegetation cover trend between 2007 and 2008 years was calculated. Results show quite a good re-growth rate along the researched route corridor.
Discussions and conclusions
The acquired results give a base for analysis of vegetation cover status during 2007 and 2008 and the trend between two years present overall statistics of vegetation trend between two years. Vegetation cover percentage identified based on satellite images allow researchers to investigate and identify areas with increase of vegetation cover, with weak trend for increase of vegetation cover or negative trend expressed in decrease of vegetation cover. Moreover this methodology enables to determine how remotely sensed data correlates with ground-truth data and accurately determines which one of the sources has discrepancies in originally measured quality (Bayramov E. R. 2005). For immense territories it is inevitable to conduct similar analysis for planning of human intervention activities, improvement of vegetation cover and for reporting process to present overall trend. Similar environmental management systems can play a significant role in improving the disturbed nature caused by human construction activities.
References
1. Erdas (2000): ERDAS Imagine Field Guide. ERDAS Inc.: Atlanta.
2. Emil R. Bayramov, Ramiz Mamedov, Coastal Landscape and Environmental Planning of the Azerbaijan Nature Protected Areas Affected by the Caspian Sea Level Fluctuation using Geographical Information Systems and Remote Sensing// Proceedings of IV International Scientific Congress “GEO-Siberia -2008”, 22-24 April 2008, Novosibirsk, Russian Federation, p. 191-198.
3. Emil. R. Bayramov, Research of the New GIS Method for the Prediction of Caspian Sea Fluctuation Impacts on the Protected Areas Located along Coastal Line for the Environmental Monitoring Purposes // Proceedings of III International Scientific Congress “GEO-Sibiria-2007”, 25 - 27 April 2007, Volume 3: Earth Remote Sensing and Photogrammetry, Environmental Monitoring, and Geoecology, Siberian State Academy of Geodesy (SSGA), Novosibirsk, Russian Federation, p. 81 - 84.
4. Bayramov E. R., Bayramov R.V (2004) Preparation of Orthophotos from IKONOS Imagery for Cadastre Base Mapping of Nakhchevan Autonomous Republic Territory, In: Proceedings of ISPRS XXXV Congress, 12-23 July 2004 Turkey, Istanbul, p. 21-22.
5. Bayramov E. R. Comparison of Object-based and Pixel-based Classification Methods Based on High Resolution Satellite Imagery, In: Proceedings of the Second International Conference, EARTH FROM SPACE - THE MOST EFFECTIVE SOLUTIONS, 30 Nov - 2Dec 2005, Russian Federation, Moscow, p. 154.
Background of authors:
Authors of this article have ten-year work experience in the area of GIS and remote sensing applied in oil and gas industry, planning, construction, environmental monitoring, land administration, cadastre, bio-restoration, corporate application development etc. working in Azerbaijan and abroad.
Contact Information of Authors:
Emil Rafael Bayramov, email: emil.bayramov@yahoo.com, bayre3@bp.com Emin Akif Hamidov, email: hamidoea@bp.com Rafael Veli Bayramov, email: rafael_karta@mail.ru
© Эмиль Р. Байрамов, Рафаэль В. Байрамов, Эмин А. Хамидов, 2009
