Analysis of advantage on Radar Remote Sensing for agricultural application

Wei Zhang Tailai Yan Shucheng You
Remote Sensing Center of China Agricultrual University
Beijing 100094, China
Tel: 86-010-62893508
[email protected]
Abstract
Because of the all-weathers and whole-day-long monitoring capabilities, radar remote sensing have many advantages on the application of agriculture. It is necessary to objectively and reasonably have an analysis of the advantage and shortcoming of Radar RS for agricultural application. There are quite a few unknown fields on the effects of microwave upon crops or plants,it is a difficult task to mine more agricultural information from a Radar RS image.

The image processing of radar RS, to develop more suitable software for the radar RS , and the foundational research are needed to pay more attention to. The radar RS has a very bright prospects future.

1. Introduction
Scientists and officials of the government in agricultural section pay more and more attention to the radar RS, which is because of the reason for radar RS not only have the all-weathers and whole-day-long monitoring capabilities, but also have many other advantages on the application of agriculture. It is necessary to objectively and reasonably have an analysis of the advantage and shortcoming of Radar RS for agricultural application in order to push the development of radar RS forward. In fact, agriculture is a very large and complicated field. It is not enough to only use the conventional RS, because of its shortcomings, some of which are crucial, for instance, in the cloudy weather conditions, it can not get the image for the surveying work by using the conventional satellite RS. The cloudy weather, in most eastern part of China, the monsoon weather areas, often happens from May to August, the time when the crops growing and needed to be monitored by RS. It is already the uniform understanding that radar RS has the unreplaceable technological advantage for agriculture.

But, after all, Radar RS for agriculture is a new field. There are quite a few unknown fields on the effects of microwave upon crops or plants, even on the statistical effects. Therefore, it is a difficult task to mine more agricultural information from a Radar RS image.

Chinese scientists and technologists on RS application for agriculture have done a lot of works on radar RS for agriculture, including land use monitoring, status monitoring for crop growth, flood and drought monitoring, and the estimation of crop yield.

2. Comparing Radar Rs with Conventional RS
The Table1 shows the differences between the Radar RS and conventional RS. The detailed explanation of its content is given as below:

• In regard to resource of electric-and- magnetic wave: Radar RS uses artificial wave (compared with sun light), so it is called the active RS. Conventional RS uses natural wave, so it is called the passive RS. This means RS technologists are able to select the suitable wave- band and polarization model based on the monitoring objects by using Radar RS. But the noise of radar wave is definitely larger than natural wave.
• With regard to wavelength: Radar wave is a microwave, the wavelength range is from 0.3cm to 100cm. Within the range, the majority of surfaces of common objects are at the critical condition, i.e., some are rough surfaces, and other ones are mirror surfaces, according to Rayligh Criteria:

  Where
  h is the roughness,
  l is the wavelength,
  b is depression angle.

  For the radar wave (microwave), the surface of water-body, road surface, leaves of plant and so on, are considered as mirror surfaces. And the surfaces of agriculture field, tree’s canopy and so on are considered as rough surfaces. We can identify quite a few kinds of objects on radar RS image by using the effect of roughness.

  On grayness of pixel depends on, for radar RS, not only roughness, but also back-scattering coefficient of target in stead of only depending on Albedo of target for conventional RS. It makes the image of radar RS more complicated, and interpretation of the image becomes more difficult.

• On color composition model of image: for conventional RS, the color composition model is false-color. It means by the model the color of pixel is closed to the condition or status of responding plant growing, which is benefit to monitoring crops and land-use. But for radar RS, the color composition model is pseudo-color. It means the color is totally not related to the condition of plant growing.

The image data give expression to body message for radar RS, but for conventional RS to surface message of target. It is because microwave has the penetrating power, but the visible or infrared light has not. This causes the increase of interpretation difficulties of radar RS imageries.

Generally speaking, radar RS data have got more information, but more difficult to interpret the image. It needs more advanced technology to process image. And it is also hopefuly to get more information on agricultural field.
Table. To compare radar RS with conventional RS

3. Analysis of Advantage and Shortcoming of Radar RS
Conventional RS has two aspects, so does radar RS as well. We discuss the advantage and shortcoming of radar RS on the following sections for the agricultural applications.

3.1. Land Use Survey
Survey of land use is the basic work on the territories and agriculture management at any scale, large or small. Image of radar RS, both spaceborne and airborne, has the following advantages:

• Easy to identify the small elevation changing of land from the image, e.g., ditch, dam, small ridge, riverside etc., even size of the objectives is smaller than the resolution of a pixel. It is the effective of multiple-surface angle reflection. This phenomenon can not be found in the image of conventional RS. Because there exists not a mirror-surface and any multiple-surface angle reflecting body on the earth for visible or infrared light.
• Easy to identify the smooth surface, e.g., water body, bare land, roads, railway, etc., which show the darken spots or lines in imageries, because of the mirror reflectiond. This phenomenon can not be found in the image of conventional RS, either.
• Industry areas, urban districts of the city, even the villages , have got apparent characters in the image of radar RS, which show itself apparently brightness. Because the surfaces of buildings form the multiple-surface angle reflecting body.

But, it should be mentioned that geometric correction is not easy because of serious top-shift and layover at range direction in the image of radar RS, especaily in. Mixed pixel is another difficult problem in RS image processing, and the problem of mixed pixel is more serious in the image of radar RS. This problem is more serious at the case of the hill or mountain areas. Some times, it influences the correct interpretation of radar image in the aspect of land use.

3.2. Monitoring Status of Crop Growing.
Theoretically speaking, there exists not any advantage by using radar RS to monitor the status of crop growing. Because the micro wave-band does not cover the range of the wave spectrum of visible and infrared light, i.e. the light of photosynthesis. Monitoring status of crop growing has to use the indirectly measuring. In fact, radar RS uses the measuring moisture content of crop for monitoring the status of crop growing .At some case, the technologists have to use radar RS for collecting information of crop growing. Because, during the season of crop growing, it often remains, at least it is cloudy. Visible or infrared light can not penetrate cloud. It is impossible for conventional RS to monitor the status of crop growing at this conditions. But using radar RS to monitor status of crop growing has a long way to go. We have to set up a great database to store and manage huge amount of data of various backscattering coefficients with various status of each crop at every growing step. China Agricultural University has set up this kind of database which include some main crops’ backscattering coefficient data.

In addition, image composition with different polarization image is helpful to interpret the status of crop growing.

3.3. Soil Moisture Measurement
Moisture content of soil is an important information for the agricultural production. The power of microwave penetrating vegetation into the soil background makes radar RS possible to measure the soil moisture content through the covered vegetations. In theory, the backscattering coefficient of substance increases with moisture content of the substance increasing, because moisture content increasing results in dielectric coefficient increasing.

At the case of avoiding the influence of roughness of the earth, radar RS can measure the moisture content of soil. American and Chinese scientists have done some successful work in this aspect. But it was only the qualitative measurement So far the quantitatively measuring moisture content of soil by both conventional RS and radar RS has still been difficult. Because there exist too many factors to influence the measurement.

3.4. The Crops Classification
For both conventional RS and radar RS, the classification for the crops is a difficult problem. There exists only a tiny difference on spectrum of reflection or backscattering coefficients among some kinds of crops.

But radar RS has a little bit advantage than conventional RS. The effect of multiple-surface angle reflecting body is helpful to identifying some kinds of crop. Because the shape of leaves and height of stem both influence the effect. And the effect influences the texture of responding area in image of radar RS. Usually, the area responding to the crop with big leaves and high stem, e.g., maize, sunflower, etc., appears as the rough texture in the image. Reversibly, the area responding to the crop with small leaves and low stem, e.g. vegetable, rice and so on, appears as the fine texture. As the background of the rice, the water or high level of moisture, the research results, using the multi-date radar data, had shown that the radar RS could play a very successful and very important role in the rice survey work future.

4. Conclusion
Radar RS has a quite different imaging-mechanics from conventional RS. Radar RS is a kind of powerful technology for observing and monitoring the earth dynamically. It is sensitive to physical property of the targets, i.e., moisture content, electro-conductbility, geometric shape, elevation change, roughness, and so on. The sensitivity of radar RS benefits the application of agriculture, e.g., monitor of land use, measurement of soil humidity, monitoring the status of crop growing and so on.

The penetrating power of microwave makes radar RS get the wider application and more information, meanwhile, makes us have to face a lot of problems in the image processing. The working range of wavelength which radar RS operates on is not the range of working wavelength of plant photosynthesis, so radar RS has to use the indirect measurement to monitor the status of crop growing. This is an important difference of the radar RS for agricultural application compared with the other kind of remote sensing systems. The image processing of radar RS is a great Challenge for the people who work on the RS application and computer technology. It is eager to develop more suitable software for the radar RS image processing. And the foundational research on the activity mechanics of microwave to plant is also eagerly needed to pay more attention to. The progress we have taken in the radar RS development showed us a very bright prospects future.

References

• Zhang Wei, Yan Tailai, 1996, Synthetical Processing of Spaceborne SAR and TM Image for Extracting the Crop Information from Sar Image Data, , 209-214
• Zhang Wei, Yan Tailai, 1995, Surveying the Growth Status of Wheat with Multi-Polar SAR data, (SAR Application Experiment)
• Guo Huadong, 1999, SAR Remote Sensing Imagery Analysis of China
• Chen Shupeng, Tong Qingxi, Guo Huadong, Study on the Remote Sensing Mechanics
• Yan Tailai, 1994, Land Management of Land Elaci, Charles, 1987, Spaceborne Radar Remote Sensing, IEEE PRESS
• Shao Yun, Guo Huadong, 1998, Effect of Polarization and Frequency Using GlobeSAR Data On Vegetation Discrimination, Geocarto International, 109(3), 71-78