برآورد شاخص‎هاي پوشش گياهي برنج با تصاوير چندزمانه راداري و اپتيك

Estimation of Rice Vegetation Indices with Multitemporal Radar and Optic Images

قابليت ها و توانايي هاي تصويربرداري راداري در بخش تكنولوژي تصويربرداري مايكروويو چشمگير است. طبقه بندي، تشخيص و پايش محصولات زراعي به‎كمك سنجش از دور، امروزه به يكي از بخش‎هاي مهم در مديريت كشاورزي تبديل شده است. به‎سبب وجود مشكلات تصاوير اپتيك در مناطق شمالي كشور (به‎سبب وجود ابر) و ناكارآمدي روش‎هاي سنتي استفاده از باند مريي و مادون قرمز و همچنين با توجه به كوچك‎بودن اندازه قطعات شالي هاي برنج، تصاوير راداري SAR (رادار روزنه مصنوعي) با قابليت‎هاي خاص خود (نفوذپذيري در هر شرايط آب‎وهوايي)، مي توانند جايگزين (يا مكمل) مناسبي براي برآورد شاخص پوشش هاي گياهي محصول برنج باشند. در پژوهش پيش رو با استفاده از تصاوير چندزمانه اپتيكي و راداري كه در سه مرحله نشا، داشت و برداشت در منطقه بهشهرِ استان مازندران انجام گرفت، به بررسي و مقايسه پنج شاخص پوشش گياهي محصول برنج در تصاوير اپتيك لندست با ضريب پراكندگي راداري ماهواره رادارست-1، در پلاريزاسيون HH پرداخته شده است. در اين پژوهش، يك مدل رياضي رگرسيون خطي با ضريب همبستگي ارايه شد و اين نتيجه به‎دست آمد كه شاخص پوشش گياهي NDVI با ضريب همبستگي 92/0 و شاخص SR با ضريب همبستگي 86/0، به‎ترتيب داراي بالاترين ضريب همبستگي با ضريب پراكندگي راداري هستند.

Introduction Due to capabilities of imaging radar, there has been an enormous surge of interest in microwave imaging technology. Unlike optical imaging, understanding the theoretical underpinnings of imaging radar can be challenging, particularly when new to the field. The technology is relatively complicated, and understanding the interaction of the incident microwave energy with the landscape to form an image has a degree of complexity well beyond that normally encountered in optical imaging. The aim of this paper is to assess the use of RADARSAT data for estimation of rice vegetation indices. The radar backscatter coefficient ?^°of rice fields appears to have a significant temporal variation. Due to weather conditions in the north of Iran, microwave sensors can be more effective in monitoring rice growth than optical sensors, since a longer wavelength electromagnetic wave is less affected by clouds and precipitation events. The backscattering measurements of rice-growing areas have already been acquired using satellite synthetic aperture radars. Time series RADARSAT fine beam mode data was acquired from May till August 1998 for seashore of Behshahr, Behshahr seashore of Mazandaran province to assess and monitor rice crop from the space. Vegetation indices (VIs) are among the oldest tools in remote sensing studies. Although many variations exist, most of them are coefficients of the reflection of light in the red and NIR ranges of the electromagnetic spectrum to separate the landscape into water, soil, and vegetation. To determine the density of green on a patch of land, researchers must observe the distinct colors (wavelengths) of visible and near-infrared sunlight reflected by the plants. As can be seen through a prism, many different wavelengths make up the spectrum of sunlight. When sunlight strikes objects, certain wavelengths of this spectrum are absorbed and other wavelengths are reflected. The pigment in plant leaves, chlorophyll, strongly absorbs visible light (from 0.4 to 0.7 µm) for use in photosynthesis. The cell structure of the leaves, on the other hand, strongly reflects near-infrared light (from 0.7 to 1.1 µm). The more leaves a plant has, the more these wavelengths of light are affected. Theoretical analyses and field studies have shown that VIs are near-linearly related to photo synthetically active radiation absorbed by a plant canopy, and therefore to light-dependent physiological processes, such as photosynthesis, occurring in the upper canopy. Practical studies have used time-series VIs to measure primary production and evapotranspiration. In this paper 5 VIs was investigated and compared with radar backscatter coefficient ?^°of rice and made a mathematical linear regression model with the correlation coefficients for estimation VIs from RADAR images. Methodology This research is a descriptive-analytical study based on acquired data and statistical methods. The following stages and procedures are to be considered: Pre-processing stage; included: 1. Co registration 2. Calibration (speckle reduction) 3. ?^°and B^° extraction from time series of RADAR images. Atmospheric correction of optical Landsat images with FLAASH (MODTRAN4) module. processing stage; included: 1. Converting DN into radiance and reflectance coefficients in optical bands of Landsat images 2. Generating NDVI, DVI, IPVI, SR, RDVI, vegetation indices from three Landsat images (red and near infrared bands). 3. Extraction of statistical parameters in 10 test site same as RADAR images. Calculation of obtained statistical data with MATLAB software and creation of linear regression equations and correlation coefficients. Results and Discussion To further explore the relationship between paddy growth stage and radar backscatter, mean backscatter values were calculated for all the test fields for three different dates. The plants showed very low backscattering in early stage of plantation –12dB to –10dB. It started increasing up to –6 dB during vegetative phase of the plants, which is due to increase in height as well canopy cover. There was an increase up to –5 dB further in reproductive stage of the plants. During ripening phase, backscatter remained almost same until the field was being harvested. This is due to not much change in plant growth during the ripening period. All considered VIs in this research shows increasing in reflectance proportional to paddy growth stage. Conclusion Because of high correlation between red and near infrared bands in optical images with chlorophylls and fresh biomass of plant (VIs) and again, high correlation between radar backscatter coefficients and biophysical parameters (content of water, canopy, height plant, plant structure), we can make a connection between those statistical parameters and create a mathematical model (simple linear regression equations) with different correlation coefficients. The results showed that the NDVI with R=0.92 has the best performance among the other four VIs. NDVI is calculated from the visible and near-infrared light reflected by vegetation. Healthy vegetation (left) absorbs most of the visible light that hits it, and reflects a large portion of the near-infrared light. Unhealthy or sparse vegetation (right) reflects more visible light and less near-infrared light. The numbers on the figure above are representative of actual values, but real vegetation is much more varied.