Two applications of the integration of GIS and Remote Sensing

Two applications of the integration of GIS and Remote Sensing

Ren Fuhu
Institute of Remote Sensing Application,
Academy of Sinica, Beijing, China

Wu Lun and Li Jing
Department of Geography, Peking Unviversity, Beijing, China

Abstract
This paper discusses the integration Remote Sensing and Geographic Information Systems through two research project: (1) monitoring water temperature and suspended sediment in Hangzhu Bay, and (2) morphotectonic research in ordos area. The research shows this kind of integration has advantages on information combination.

Introduction
Remote Sensing and Geographic Information System (GIS) are two important technology for spatial information receiving and processing, and now widely applied to many fields of resources and environment management. The integration of Remote Sensing and GIS is a trend of geographic information science, which will provide GIS sufficient information and make Remote Sensing data easy to be easy to be analyzed. The Integration is described in the following aspects.

Remote Sensing image correction
Setting up DEM from Remote Sensing data directly
Combined display:
1. 3- dimentional display of images combined with DEM
2. Overlayed thematic map on Remote Sensing image and showing results of attribute data on the image.
3. False color display of Remote Sensing image and showing results of attribute as one or two bands and significant maps as other bands.
Thematic information extraction:
1. Multi-classification of Remote Sensing images with maps
2. Determining the mixed objects in pixels
3. Extracting training areas with thematic maps.
Processing remote sensed images with GIS operations.
Processing thematic maps with image processing functions

In our work, the image processing system I2S and micro-GIS PURSIS ARE Connected together to form an powerful system (Fig. 1)

Fig. 1 Structure of the integrated System

Monitoring water temperature and suspended sediment in Hangzhou bay
Water temperature and suspended sediment are two most important indexes in ocean investigations. Different from dry land, ocean water is in continuous movement and the water indexes are changing over. time. The conventional measures by ship and hydrometric station can only get few data which is not synchronous. Remote sensed images at one fixed time can not reveal the a general moving pattern. We can get a lot of synchronous data and for combined analysis of these images by using GIS. In hangzhou bay, the whole work was divided into four connective parts (Fig.2)

Data Collection
Twenty-one AVHRR images have been received seperately in November and December of 1987 and in July and August of 1988. Dozens of distribution maps of suspended sediment were collected also for dynamic analyses. At the same time, water samples are collected in the Hangzhou Bay.
Pre-processing and Translation
1. Geting eight high bits data from digital tape of AVHRR images to maintain most information of water temperature and sediment..
2. Transforming the original images to Mercator projection to be consistent with sea map.
3. Calculating with AVHRR-CH4 and AVHRR-CH5 bye the following formula to get radiance:
  
  L = S * N + I
  where L is the radiance in corresponding channel, N and I are constants. Then calculating equivalent temperature TB with Planck's Radiance Formula, and radiance correction with the formula:
  
  T = TB + T'
  where T is the real water temperature and T, is temperature corrected due to atmospheric decreation which can be calculated from real-measured sample data.
4. Translation of AVHRR-CH1 in terms of the consistency of suspended sediment (S) with the formula:
  
  S=A+B*1n (D-L)
  where A,B,D are constants calculated with real-measured sample data.
Synthetic Calculation in PURSIS
1. Calculating the average temperature distributions in the summer and in the winter respectively.
2. Calculating the distribution of suspended sediment at six different tidal times (0,+2,+4,+6,-4,-2) with maps from remote sensing information and historical maps, then calculating the arithmetic mean of these six distribution maps.
  
  The definition of appearance frequency of high trubid water is following: consider any point A( x, y,) in water. For all N images, if A is in the high turbid water on the M images, then M is defined as the appearance frequence of high turbid water at point A, i. e :
  
  Eq.
  Calculating all points in water on the Hangzhou Bay with:
  
  Eq.
  to get the distribution map of the appearance frequency of the high turbid water.
Fig. 2 Working procedures
Result Analysis
One the water temperature maps, we can get the dynamic characterristics of water temperature in the seasons of summer and winter. In the summer, the temperature within the bay is higher than the one outside the bay, and the position of the highest temperature is at the top of the bay, the isolines area of south-north. On the contrary, the temperature of inner bay is lower than outer bay in winter. Two cold water bodies ware discovered, one is at the top of the bay, the other is at Nanhuizhui. Warmer water comes from zhoushan islands and intserts in hangzhou bay forming a sharp curve front.

From the tubidity maps, the general distributive pattern of suspended sediment and its movement with tide can be seen clearly at a glance in summer, the high turbid water (S>400 mg/1) forms a band water mass across the bay. the highest S (S> 625 mg/1) are in the east of Nanhui and the north of Andong. S is low outside the bay. In the bay, the region of lowest S is from Zhapu to Janshan, where should be one of the best place to build harbor in the hangzhou Bay.

The position of high turbid water changes periodically with tide. at high tide (tidal time 0), it is located in the most inside; Then it moves to outside with ebb tide. At tidal time +2, it is near the Tanxu; At low tide (tidal time+6), it flows from the Tanxu to the Andon. The high turbid water near the Nanhui has developed obviously, and extends to the Tanxu at tidal time -4. At tidal time-2, it already removes to inside. In the average map (Fig. 3) we can see clearly the centrical position of the high turbid water and its affecting range.

Fig. 3 Probability distribution map of maximum turbidity of 1976 - 1988
in Hangzhou bay of China.
The application on morphotectonic research in ordos area
Remote Sensing images contain a lot of information about landform. The integration of thematic maps and geophysical data ( involved in composition of surface, characteristic of deep-crustal structure, old and new tectonic movement etc.) with remote sensing information is very important and significant for Morphotectonic analysis.

In this study, we first pre-process grid images dealing with geophysical and geodetic measurement data based on PURSIS, including crustal thickness (as result of tectonic movement within Mesozoic era or earlier ) Neotectonic movement since Neogene and modern crustal deformaion. We then process NOAA-AVHRR images by the supprot of I2S, and combine GIS with the images to study the law of tectonic movement and landform, with an enhancement of geomorphic interpretation.

Data Gathering and pre-processing
1. Remote sensing image processing : adopting CCT tape of NOAA-AVHRR (0ct. 30, 1989, CH1-CH4, to get 788x500 digital image in Ordos area ( ranges E105-E111 and N34-N41. 40) after projection translation and digital mosaic.
2. Producing digital ( raster) images o crustal thickness, ground deformation and quartnary deposition thickness: Inputing each isogram and interpolating.
3. Producing DEM: Inputing eight topography maps of the area and combine them together, then, interpolating of DEM.
4. Calculating of vertical Neotectonic movement: Calaculating landform envelope surface, cutting down the thickness of Q deposition to get digital map of Neogene plantation the ( which widely exists in this area ), whose elevation represents of vertical Neotectonic movement.
Integrating of Remote Sensing Image with Thematic Information
Supported by I2S, false-color images with each composition schemes of different bands and themes are displayed in (Table 1).

According to Chromatics, the mixed image has the following characteristics : When composite elements are identical, it displays white (the three elements are all strong) or black -grey ( all weak), and when one of the three is not the same (Strong or weak) as the others, it displays primary or complementary colors.

In our experiment with thematic maps, generally all elements are identical and most area are white or grey. It shows that, the areas with thin crust and weak uplifting or subsidence movement, are usually low basin or plain in landform, meanwhile, the mountain and plateau area correspond with strong uplifting movement and thick crust.

For those remote sensing images joined with thermatic information, the faults zone interpretation becomes much easier and more accurate.

Robert and Sobel processing with the former mixed images can extract sharp grade zone. The active toctonic zones in this area are displayed (Fig. 4), those had proved by Earthquake distribution. This method is helpful for earthquake research, construction foundation survey and resources development ( such as ground water, oil) et al.

Fig. 4 Active tectonic zones in ordos area (by Robert translation
from mixed images of AVHRR and thematic maps)

Table 1. Integrating remote sensed image with thematic information

Scheme	Content		Geomorphic significance
1	Red	crust thickness	Relation among landform, Neotectonic movement and ground deformation.
	green	DEM
	Blue	Neotec Movement
2	Red	crust thickness	Tectonic characteristic of different Geological period, and its spatial distribution.
	green	Neotec. Movement
	Blue	ground deformation
3	red	crust thickness	Relation among landform and crustal structure, and the recent trend of landform change
	green	DEM
	blue	ground deformation
4	red	Neotech. Movement	Relation between landform and Neotectonic movement
	green	DEM
	blue	ground deformation
5	red	DEM	Information enhancement for Geomorphci interpretation
	green	AVHRR-CH3
	blue	AVHRR-CH1
6	red	DEM	Information enhancement for Neotectonic interpretation
	green	AVHRR-CH4
	blue	AVHRR-CH1
7	red	crust thickness	Information integration of deep -part and surface of crust
	green	DEM
	blue	AVHRR-CH1
8	red	crust thickness	enhancement tectonic information for fault interpretation
	green	Neotec. Movement
	blue	AVHRR-CH1
9	red	Neotec. Movement	Charactersistic and tendency of new and modern tectonic movement, and its expression on earth surface
	green	ground deformation
	blue	AVHRR-CH1