Sungwon Choi

Long-Term Variability of Surface Albedo and Its Correlation with Climatic Variables over Antarctica

The cryosphere is an essential part of the earth system for understanding climate change. Components of the cryosphere, such as ice sheets and sea ice, are generally decreasing over time. However, previous studies have indicated differing trends between the Antarctic and the Arctic. The South Pole also shows internal differences in trends. These phenomena indicate the importance of continuous observation of the Polar Regions. Albedo is a main indicator for analyzing Antarctic climate change and is an important variable with regard to the radiation budget because it can provide positive feedback on polar warming and is related to net radiation and atmospheric heating in the mainly snow- and ice-covered Antarctic. Therefore, in this study, we analyzed long-term temporal and spatial variability of albedo and investigated the interrelationships between albedo and climatic variables over Antarctica. We used broadband surface albedo data from the Satellite Application Facility on Climate Monitoring and data for several climatic variables such as temperature and Antarctic oscillation index (AAO) during the period of 1983 to 2009. Time series analysis and correlation analysis were performed through linear regression using albedo and climatic variables. The results of this research indicated that albedo shows two trends, west trend and an east trend, over Antarctica. Most of the western side of Antarctica showed a negative trend of albedo (about −0.0007 to −0.0015 year−1), but the other side showed a positive trend (about 0.0006 year−1). In addition, albedo and surface temperature had a negative correlation, but this relationship was weaker in west Antarctica than in east Antarctica. The correlation between albedo and AAO revealed different relationships in the two regions; west Antarctica had a negative correlation and east Antarctica showed a positive correlation. In addition, the correlation between albedo and AAO was weaker in the west. This suggests that the eastern area is influenced by the atmosphere, but that the western area is influenced more strongly by other factors.

New Approach for Snow Cover Detection through Spectral Pattern Recognition with MODIS Data

Snow cover plays an important role in climate and hydrology, at both global and regional scales. Most previous studies have used static threshold techniques to detect snow cover, which can lead to errors such as misclassification of snow and clouds, because the reflectance of snow cover exhibits variability and is affected by several factors. Therefore, we present a simple new algorithm for mapping snow cover from Moderate Resolution Imaging Spectroradiometer (MODIS) data using dynamic wavelength warping (DWW), which is based on dynamic time warping (DTW). DTW is a pattern recognition technique that is widely used in various fields such as human action recognition, anomaly detection, and clustering. Before performing DWW, we constructed 49 snow reflectance spectral libraries as reference data for various solar zenith angle and digital elevation model conditions using approximately 1.6 million sampled data. To verify the algorithm, we compared our results with the MODIS swath snow cover product (MOD10_L2). Producer’s accuracy, user’s accuracy, and overall accuracy values were 92.92%, 78.41%, and 92.24%, respectively, indicating good overall classification accuracy. The proposed algorithm is more useful for discriminating between snow cover and clouds than threshold techniques in some areas, such as those with a high viewing zenith angle.

Retrieval of background surface reflectance with BRD components from pre-running BRDF

Many countries try to launch satellite to observe the Earth surface. As important of surface remote sensing is increased, the reflectance of surface is a core parameter of the ground climate. But observing the reflectance of surface by satellite have weakness such as temporal resolution and being affected by view or solar angles. The bidirectional effects of the surface reflectance may make many noises to the time series. These noises can lead to make errors when determining surface reflectance. To correct bidirectional error of surface reflectance, using correction model for normalized the sensor data is necessary. A Bidirectional Reflectance Distribution Function (BRDF) is making accuracy higher method to correct scattering (Isotropic scattering, Geometric scattering, Volumetric scattering). To correct bidirectional error of surface reflectance, BRDF was used in this study. To correct bidirectional error of surface reflectance, we apply Bidirectional Reflectance Distribution Function (BRDF) to retrieve surface reflectance. And we apply 2 steps for retrieving Background Surface Reflectance (BSR). The first step is retrieving Bidirectional Reflectance Distribution (BRD) coefficients. Before retrieving BSR, we did pre-running BRDF to retrieve BRD coefficients to correct scatterings (Isotropic scattering, Geometric scattering, Volumetric scattering). In pre-running BRDF, we apply BRDF with observed surface reflectance of SPOT/VEGETATION (VGT-S1) and angular data to get BRD coefficients for calculating scattering. After that, we apply BRDF again in the opposite direction with BRD coefficients and angular data to retrieve BSR as a second step. As a result, BSR has very similar reflectance to one of VGT-S1. And reflectance in BSR is shown adequate. The highest reflectance of BSR is not over 0.4μm in blue channel, 0.45μm in red channel, 0.55μm in NIR channel. And for validation we compare reflectance of clear sky pixel from SPOT/VGT status map data. As a result of comparing BSR with VGT-S1, bias is from 0.0116 to 0.0158 and RMSE is from 0.0459 to 0.0545. They are very reasonable results, so we confirm that BSR is similar to VGT-S1. And weakness of this study is missing pixel in BSR which are observed less time to retrieve BRD components. If missing pixels are filled, BSR is better to retrieve surface products with more accuracy. And we think that after filling the missing pixel and being more accurate, it can be useful data to retrieve surface product which made by surface reflectance like cloud masking and retrieving aerosol.

Snow cover detection algorithm using dynamic time warping method and reflectances of MODIS solar spectrum channels

Snow cover is biggest single component of cryosphere. The Snow is covering the ground in the Northern Hemisphere approximately 50% in winter season and is one of climate factors that affects Earth’s energy budget because it has higher reflectance than other land types. Also, snow cover has an important role about hydrological modeling and water resource management. For this reason, accurate detection of snow cover acts as an essential element for regional water resource management. Snow cover detection using satellite-based data have some advantages such as obtaining wide spatial range data and time-series observations periodically. In the case of snow cover detection using satellite data, the discrimination of snow and cloud is very important. Typically, Misclassified cloud and snow pixel can lead directly to error factor for retrieval of satellite-based surface products. However, classification of snow and cloud is difficult because cloud and snow have similar optical characteristics and are composed of water or ice. But cloud and snow has different reflectance in 1.5 ~ 1.7 μm wavelength because cloud has lower grain size and moisture content than snow. So, cloud and snow shows difference reflectance patterns change according to wavelength. Therefore, in this study, we perform algorithm for classifying snow cover and cloud with satellite-based data using Dynamic Time Warping (DTW) method which is one of commonly used pattern analysis such as speech and fingerprint recognitions and reflectance spectral library of snow and cloud. Reflectance spectral library is constructed in advance using MOD21km (MODIS Level1 swath 1km) data that their reflectance is six channels including 3 (0.466μm), 4 (0.554μm), 1 (0.647μm), 2 (0.857μm), 26 (1.382μm) and 6 (1.629μm). We validate our result using MODIS RGB image and MOD10 L2 swath (MODIS swath snow cover product). And we use PA (Producer’s Accuracy), UA (User’s Accuracy) and CI (Comparison Index) as validation criteria. The result of our study detect as snow cover in the several regions which are did not detected as snow in MOD10 L2 and detected as snow cover in MODIS RGB image. The result of our study can improve accuracy of other surface product such as land surface reflectance and land surface emissivity. Also it can use input data of hydrological modeling.

Long-term variability of Total PrecipitableWater using a MODIS over Korea

Abstract : Water vapor leading various scale of atmospheric circulation and accounting for about60% of the naturally occurring warming effect is important climate variables. Using the Total PrecipitableWater (TPW) from Moderate Resolution Imaging Spectroradiometer (MODIS) operating on both Terraand Aqua, we study long-term Variation of TPW and define relationship among TPW and climaticparameters such as temperature and precipitation to quantitatively demonstrate the impact on climatechange over East Asia focusing on the Korea peninsula. In this study, we used linear regression analysisto detect the correlation of TPW and temperature/precipitation and harmonic analysis to analyzechangeable aspects of periodic characteristics. A result of analysis using linear regression analysis betweenTPW and climate elements, TPW shows a high determination coefficient (R 2 ) with temperature andprecipitation (determination coefficient between TPW and temperature: 0.94, determination coefficientbetween TPW anomaly and temperature anomaly: 0.8, determination coefficient between TPW andprecipitation: 0.73, determination coefficient between TPW anomaly and precipitation anomaly: 0.69). Aresult of harmonic analysis of TPW and precipitation of two-year to five-year cycle, amplitude contributionratio of 3.5-year cycle are much higher and two phases are similar in 3.5-year cycle.Key Words : MODIS, TPW, Anomaly, Harmonic analysis요약 :