Description: The discrete wavelet transform (DWT)-based Set Partitioning in Hierarchical Trees (SPIHT) algorithm is widely used in many image compression systems. The time-consuming computation of the 9/7 discrete wavelet decomposition is usually the bottleneck of these systems. In order to perform real-time Reed-Solomon channel decoding and SPIHT+DWT source decoding on a massive bit stream of compressed images continuously down-linked from the satellite, we propose a novel graphic processing unit (GPU)-accelerated decoding system. In this system the GPU is used to compute the time-consuming inverse DWT, while multiple CPU threads are run in parallel for the remaining part of the system. Both CPU and GPU parts were carefully designed to have approximately the same processing speed to obtain the maximum throughput via a novel pipeline structure for processing continuous satellite images. As part of the SPIHT decoding system, the GPU-based inverse DWT is about 158 times faster than its CPU counterpart. Through the pipelined CPU and GPU heterogeneous computing, the entire decoding system approaches a speedup of 83x as compared to its single-threaded CPU counterpart. The proposed channel and source decoding system is able to decompress 1024x1024 satellite images at a speed of 90 frames per second.

Description: In this paper, we develop a novel Graphics Processing Unit (GPU)-based high-performance Radiative Transfer Model (RTM) for the Infrared Atmospheric Sounding Interferometer (IASI) launched in 2006 onboard the first European meteorological polar-orbiting satellites, METOP-A. The proposed GPU RTM processes more than one profile at a time in order to gain a significant speedup compared to the case of processing just one profile at a time. The radiative transfer model performance in operational numerical weather prediction systems nowadays still limits the number of channels they can use in hyperspectral sounders to only a few hundreds. To take the full advantage of such high resolution infrared observations, a computationally efficient radiative transfer model is needed. Our GPU-based IASI radiative transfer model is developed to run on a low-cost personal supercomputer with 4 NVIDIA Tesla C1060 GPUs with total 960 cores, delivering near 4 TFlops theoretical peak performance. The model exhibited linear scaling with the number of graphics processing units. Computing 10 IASI radiance spectra simultaneously on a GPU, we reached 763x speedup for 1 GPU and 3024x speedup for all 4 GPUs, both with respect to the original single-threaded Fortran CPU code. The significant 3024x speedup means that the proposed GPU-based high-performance forward model is able to compute one day's amount of 1,296,000 IASI spectra within 6 minutes, whereas the original CPU-based version will impractically take more than 10 days. The GPU-based high-performance IASI radiative transfer model is suitable for the assimilation of the IASI radiance observations into the operational numerical weather forecast model.

Description: For the large-volume ultraspectral sounder data, compression is desirable to save storage space and transmission time. To retrieve the geophysical paramters without losing precision the ultraspectral sounder data compression has to be lossless. Recently there is a boom on the use of graphic processor units (GPU) for speedup of scientific computations. By identifying the time dominant portions of the code that can be executed in parallel, significant speedup can be achieved by using GPU. Predictive partitioned vector quantization (PPVQ) has been proven to be an effective lossless compression scheme for ultraspectral sounder data. It consists of linear prediction, bit depth partitioning, vector quantization, and entropy coding. Two most time consuming stages of linear prediction and vector quantization are chosen for GPU-based implementation. By exploiting the data parallel characteristics of these two stages, a spatial division design shows a speedup of 72x in our four-GPU-based implementation of the PPVQ compression scheme.

Description: Harmful algal blooms (HABs) pose an enormous threat to the U.S. marine habitation and economy in the coastal waters. Federal and state coastal administrators have been devising a state-of-the-art monitoring and forecasting system for these HAB events. The efficacy of a monitoring and forecasting system relies on the performance of HAB detection. We propose a machine learning based spatio-temporal data mining approach for the detection of HAB events in the region of the Gulf of Mexico. In this study, a spatio-temporal cubical neighborhood around the training sample is introduced to retrieve relevant spectral information of both HAB and non-HAB classes. The feature relevance is studied through mutual information criterion to understand the important features in classifying HABs from non-HABs. Kernel based support vector machine is used as a classifier in the detection of HABs. This approach gives a significant performance improvement by reducing the false alarm rate. Further, with the achieved classification accuracy, the seasonal variations and sequential occurrence of algal blooms are predicted from spatio-temporal datasets. New variability visualization is introduced to illustrate the dynamic behavior of HABs across space and time.

Description: Radiative transfer modelling of high resolution infrared (or microwave) spectra still represents a major challenge for the processing of atmospheric remote sensing data despite significant advances in the numerical techniques utilized in line-by-line modelling by, e.g., optimized Voigt function algorithms or multigrid approaches. Special purpose computing hardware such as Field Programmable Gate Arrays (FPGAs) can be used to cope with the dramatic increase of data quality and quantity. Utilizing a highly optimized implementation of an uniform rational function approximation of the Voigt function, the molecular absorption cross section computation-representing the most compute intensive part of radiative transfer codes-has been realized on FPGA. Design and implementation of the FPGA coprocessor is presented along with first performance tests and an outlook for the ongoing further development.

Description: Today, the impact of irrigation on land surface attributes is not well documented. Existing studies are focused on meteorological observations and modeling. This paper aims to validate whether satellite observations are capable of detecting the impact of irrigation on land surface attributes. We used satellite observations to estimate land surface parameters, including surface albedo, land surface temperature (LST), Normalized Difference Vegetation Index (NDVI), soil moisture (SM), and evapotranspiration (ET) for the irrigation agriculture system. This study has two objectives: first, determining if satellite observations can detect the irrigation impact on land surface attributes; and second, to quantitatively evaluate the land surface attributes under different irrigation intensities. Results show that from 2000 to 2008 the land surface parameters of irrigated areas had obvious intra-annual variations, and highly irrigated areas feature a corresponding lower albedo and LST, higher soil moisture, NDVI and ET. This proves that satellite observations can effectively assess the irrigation impacts on land surface parameters and provide another valid method for determining the impact of irrigation on the local surface climate-especially in those regions where direct observations are limited or obscured by other factors, such as urbanization in China.

Description: Ground-penetrating radar (GPR) measurement and its interpretation/analysis are challenging when soil is heterogeneous. Soil heterogeneity causes unwanted reflections (i.e., clutter) that disturb reflections from objects of interest. Thorough investigations on soil heterogeneity and clutter are important in order to understand the influence on GPR and assess the performance. In order to observe the influence of heterogeneous soil, an irrigation test was carried out and GPR data were collected after the irrigation and while the distribution of soil water content varied. The correlation length and variability of the dielectric constant of soil were determined by geostatistical analyses of the GPR data. These two parameters were built into a simple model and the Mie solution was theoretically calculated. From this, the power of the backscattered field due to soil heterogeneity was modeled. The results were in agreement with the power of the clutter extracted from the GPR data. Therefore, clutter can be predicted from soil heterogeneity with a simple model using the Mie solution. Furthermore, the result exhibits that scattering by heterogeneous soil is dominated by Mie scattering, rather than Rayleigh scattering, in the studied frequency range.