Search

Weather satellites are vital tools for monitoring the global environment. They provide data that is used to deliver essential predictions and warnings, and to save lives. Space agencies have for years recognized the importance of sharing remotely sensed data for weather analysis and forecasting, climate analysis, and monitoring hazards worldwide. As a result, they operate under a policy of freely shared data. The U.S. Joint Polar Satellite System (JPSS), a contributor to the global observing system, provides key observables that are crucial to obtaining continuity, global coverage, and filling data gaps. JPSS collaborates with national and international partners through the WMO and engages with multilateral organizations such as CGMS, GEO, and CEOS to develop requirements, establish best practices for combining measurements from multiple satellite sensors, and develop capabilities that enable communities worldwide to develop local solutions to address the challenges related to global atmospheric processes and their complex interactions.

In this paper, we describe how researchers and weather forecasters work together to make satellite sounding data sets more useful in severe weather forecasting applications through participation in National Oceanic and Atmospheric Administration (NOAA)’s Hazardous Weather Testbed (HWT) and JPSS Proving Ground and Risk Reduction (PGRR) program. The HWT provides a forum for collaboration to improve products ahead of widespread operational deployment. We found that the utilization of the NOAA-Unique Combined Atmospheric Processing System (NUCAPS) soundings was improved when the product developer and forecaster directly communicated to overcome misunderstandings and to refine user requirements. Here we share our adaptive strategy for (1) assessing when and where NUCAPS soundings improved operational forecasts by using real, convective case studies and (2) working to increase NUCAPS utilization by improving existing products through direct, face-to-face interaction. Our goal is to discuss the lessons we learned and to share both our successes and challenges working with the weather forecasting community in designing, refining, and promoting novel products. We foresee that our experience in the NUCAPS product development life cycle may be relevant to other communities who can then build on these strategies to transition their products from research to operations (and operations back to research) within the satellite meteorological community.

An Earth-observing remote sensing instrument is used to collect information about the physical environment within its instantaneous-field-of-view and is often placed aboard a suborbital or satellite platform for maximal spatial coverage. Remote sensing inversion techniques can extract valuable meteorological parameters that are subsequently passed through weather models for research and forecasting. One of the several microphysics packages for clouds and precipitation in weather models is WRF Single Moment 6-class (WSM6) scheme, and it is now widely used. With the advancement in Graphics Processing Units (GPUs), implementation of a fast and parallel WSM6 scheme is achievable. This paper describes a massively parallel GPU design of the WSM6 scheme. The performance is compared to a CPU implementation running on Intel Xeon E5-2603 at 1.8GHz. Our implementation shows a speedup of 216× using a single NVIDIA K40 GPU as compared to its CPU counterpart running on one CPU core, whereas the speedup for one CPU socket (4 cores) with respect to one CPU core is only 3.7×. We develop a massively parallel GPU design of the WRF WSM6 microphysics scheme.NVIDIA CUDA C programming language is used to implement the WSM6 scheme.We show a speedup of 216× using a single NVIDIA K40 GPU.The speedup for one CPU socket (4 cores) with respect to one CPU core is only 3.7×.

Modern weather satellites provide more detailed observations of cloud and precipitation processes. To harness these observations for better satellite data assimilations, a cloud-resolving model, known as the Goddard Cumulus Ensemble (GCE) model, was developed and used by the Goddard Satellite Data Simulator Unit (G-SDSU). The GCE model has also been incorporated as part of the widely used weather research and forecasting (WRF) model. The computation of the cloud-resolving GCE model is time-consuming. This paper details our massively parallel design of GPU-based WRF GCE scheme. With one NVIDIA Tesla K40 GPU, the GPU-based GCE scheme achieves a speedup of ${\bf 361} \times $ as compared to its original Fortran counterpart running on one CPU core, whereas the speedup for one CPU socket (four cores) with respect to one CPU core is only ${\bf 3}.{\bf 9} \times $ .

In recent years, graphics processing units (GPUs) have emerged as a low-cost, low-power and a very high performance alternative to conventional central processing units (CPUs). The latest GPUs offer a speedup of two-to-three orders of magnitude over CPU for various science and engineering applications. The Weather Research and Forecasting (WRF) model is the latest-generation numerical weather prediction model. It has been designed to serve both operational forecasting and atmospheric research needs. It proves useful for a broad spectrum of applications for domain scales ranging from meters to hundreds of kilometers. WRF computes an approximate solution to the differential equations which govern the air motion of the whole atmosphere. Kessler microphysics module in WRF is a simple warm cloud scheme that includes water vapor, cloud water and rain. Microphysics processes which are modeled are rain production, fall and evaporation. The accretion and auto-conversion of cloud water processes are also included along with the production of cloud water from condensation. In this paper, we develop an efficient WRF Kessler microphysics scheme which runs on Graphics Processing Units (GPUs) using the NVIDIA Compute Unified Device Architecture (CUDA). The GPU-based implementation of Kessler microphysics scheme achieves a significant speedup of 70x over its CPU based single-threaded counterpart. When a 4 GPU system is used, we achieve an overall speedup of 132x as compared to the single thread CPU version.

In summary, the JPSS Proving Ground Program has been a huge success. The innovative use of S-NPP data and the full integration of the user community ensured thorough evaluation of capabilities and rapid transition from research to operations. The JPSS Program's multi-year commitment to this program allows project teams to define a successful path to assist the user community in fully integrating new JPSS capabilities.

The microwave sounding unit (MSU) prelaunch thermal-vacuum chamber test data for eight MSU flight models that flew on TIROS-N, NOAA 7 through NOAA 12, and NOAA 14 were reanalyzed using an improved calibration algorithm, originally designed for the advanced microwave sounding unit-A (AMSU-A) operations. The new calibration algorithm can automatically adjust for any channel gain shift in operation. Adoption of this calibration algorithm as the MSU calibration procedure in the recalibration project will make the data sets from the MSU and the AMSU-A more consistent. This will be useful for future blending of the climate trends generated from the MSU and AMSU-A data. A single nonlinearity parameter u, which appears in the new calibration algorithm, was obtained for each channel from analysis of the prelaunch calibration test data. A software package for implementing this new calibration algorithm was developed and applied to calculate the MSU time series for improvement of the accuracy of the climate record. Sample calculations of MSU antenna temperatures with the new calibration algorithm were made for two satellites and are compared with similar results obtained from the old MSU calibration algorithm. Significant differences are observed, and strong evidence indicates that the new calibration procedure will provide a more accurate quantification of climate trends.

The Advanced Microwave Sounding Unit-A (AMSU-A) is the first of a new generation of polar-orbiting cross-track microwave sounders operated by the National Oceanic and Atmospheric Administration. A feature of a cross-track sounder is that the measurements vary with scan angle because of the change in the optical pathlength between the earth and the satellite. This feature is called the limb effect and can be as much as 30 K. One approach to this problem is to limb adjust the measurements to a fixed view angle. This approach was used for the older series of Microwave Sounding Units. Limb adjusting is important for climate applications and regression retrieval algorithms. This paper describes and evaluates several limb adjustment procedures. The recommended procedure uses a combined physical and statistical technique. The limb adjusted measurements were compared with computed radiances from radiosondes and National Centers for Environmental Prediction models. The model error was found to be less tha...

A better understanding of the relationship between satellites observed vegetation health, and malaria epidemics could help mitigate the worldwide increase in incidence of mosquito-transmitted diseases. This research investigates last 17- years association between vegetation health (condition) index and malaria transmission in Bikaner, Rajasthan in India an arid and hot summer area. The vegetation health (condition) index, derived from a combination of Advanced Very High Resolution Radiometer (AVHRR) based Normalized Difference Vegetation Index (NDVI) and 10-μm to 11-μm thermal radiances, was designed for monitoring moisture and thermal impacts on vegetation health. We demonstrate that thermal condition is more sensitive to malaria transmission with different seasonal malaria activities. The weekly VH indices were correlated with the epidemiological data. A good correlation was found between malaria cases and Temperature Condition Index (TCI) one at least two months earlier than the malaria transmission season. Following the results of correlation analysis, Principal Component Regression (PCR) method was used to construct a model of less than 10% error to predict malaria as a function of the TCI.

Nakanishi and Niino proposed an improved MelloriVYamada (M-Y) Level-3 model (MYNN) surface for three-dimensional simulation of advection fog. The model is on the basis of large-eddy simulation and is numerically stable. The model predicts vertical profiles of mean quantities such as temperature that are in good agreement with those obtained from large-eddy simulation of a radiation fog. In this paper, we accelerate Nakanishi and Niino PBL's Surface Layer scheme in a highly parallel environment, using NVIDIA Graphics Processing Units (GPU). This GPU implementation efficiently utilizes the fine grained parallelism exhibited by the MYNN PBL scheme. The algorithm is accelerated on a low-cost personal supercomputer with over 500 CUDA cores running on a GPU. This implementation achieves a high speedup of 160x.

The Eta Ferrier cloud microphysics scheme is a sophisticated cloud microphysics module in the Weather Research and Forecasting (WRF) model. In this paper, we present the approach and the results of accelerating the Eta Ferrier microphysics scheme on NVIDIA Graphics Processing Units (GPUs). We discuss how our GPU implementation takes advantage of the parallelism in Eta Ferrier scheme, leading to a highly efficient GPU acceleration. We implement the Eta Ferrier microphysics scheme on NVidia GTX 590 GPU. Our 1-GPU implementation achieves an overall speedup of 37 as compared with a single thread CPU. Since Eta Ferrier microphysics scheme is only an intermediate module of the entire WRF model, the GPU I/O should not occur, i.e. its input data should be already available in the GPU global memory from previous modules and its output data should reside at the GPU global memory for later usage by other modules. The speedup without the host-device data transfer time is 272 with respect to its serial version running on 3.20GHz Intel® CoreTM i7 970 CPU.

The Weather Research and Forecasting (WRF) is a system of numerical weather prediction and atmospheric simulation with dual purposes for forecasting and research. The WRF software infrastructure consists of several components such as dynamic solvers and physical simulation modules. WRF includes several Land-Surface Models (LSMs). The LSMs use atmospheric information, the radiative and precipitation forcing from the surface layer scheme, the radiation scheme, and the microphysics/convective scheme all together with the lands state variables and land-surface properties, to provide heat and moisture fluxes over land and sea-ice points. The WRF 5-layer thermal diffusion simulation is an LSM based on the MM5 5-layer soil temperature model with an energy budget that includes radiation, sensible, and latent heat flux. The WRF LSMs are very suitable for massively parallel computation as there are no interactions among horizontal grid points. More and more scientific applications have adopted graphics processing units (GPUs) to accelerate the computing performance. This study demonstrates our GPU massively parallel computation efforts on the WRF 5-layer thermal diffusion scheme. Since this scheme is only an intermediate module of the entire WRF model, the I/O transfer does not involve in the intermediate process. Without data transfer, this module can achieve a speedup of 36x with one GPU and 108x with four GPUs as compared to a single threaded CPU processor. With CPU/GPU hybrid strategy, this module can accomplish a even higher speedup, ~114x with one GPU and ~240x with four GPUs. Meanwhile, we are seeking other approaches to improve the speeds.

The Weather Research and Forecast (WRF) Double Moment 5-class (WDM5) mixed ice microphysics scheme predicts mixing ratio of hydrometeors and their number concentrations for warm rain species including clouds and rain. WDM5 can be computed in parallel in the horizontal domain using a many-core GPU. In order to obtain better GPU performance, we manually rewrote the original WDM5 Fortran module into a highly parallel CUDA C program. The GPU-based implementation of WDM5 microphysics scheme on 1 GTX590 GPU achieves a significant speedup of 147× over its CPU-based single-threaded counterpart when we use asynchronous data transfer and non-coalesced memory access. More importantly, the speedup excluding the host-device data transfer time is 206× when using coalesced memory access. Since the WDM5 microphysics scheme is only an intermediate module of the entire WRF model, its input data should be already available in the GPU global memory from previous modules and its output data should reside at the GPU global memory for later usage by other modules.

The Weather Research and Forecast (WRF) model is the most widely used community weather forecast and research model in the world. There are several single moment ice microphysics models in WRF. A mixed phase for WRF Single Moment (WSM) represents the condensation, precipitation, and thermodynamic effects of latent heat release. In this paper, we will show our optimization efforts on WSM5. The processing time can be reduced from 16928 ms on CPU to 48.3 ms using General Purpose Graphics Processing Unit (GPGPU). Thus, the speedup is 350x without I/O using a single GPU. Taking I/O transfer times into account the speedup is 202x.

The Weather Research and Forecasting (WRF) model is the latest-generation numerical weather prediction model. It has been designed to serve both operational forecasting and atmospheric research needs. It proves useful for a broad spectrum of applications for scales ranging from meters to thousands of kilometers. WRF computes an approximate solution to the differential equations which govern the air motion of the whole atmosphere. Kessler microphysics module in WRF is a simple warm cloud scheme that includes water vapor, cloud water and rain. Microphysics processes which are modeled are rain production, fall and evaporation. The accretion and auto-conversion of cloud water processes are also included along with the production of cloud water from condensation. In this paper, we develop an efficient WRF Kessler microphysics scheme which runs on Graphics Processing Units (GPUs) using the NVIDIA Compute Unified Device Architecture (CUDA). The GPU-based implementation of Kessler microphysics scheme achieves a significant speedup of 70x over its CPU based single-threaded counterpart. The speedup on a GPU without host-device data transfer time is 816x. Since Kessler microphysics scheme is just an intermediate modules of the entire WRF model, the GPU I/O should not occur, i.e. its input data should be already available in the GPU global memory from previous modules and their output data should reside at the GPU global memory for later usage by other modules. Thus, the limited scaling of Kessler scheme with I/O will not be an issue once all modules have been rewritten using CUDA. High speed WRF running completely on GPUs promises more accurate forecasts in considerably less time.

The Weather Research and Forecasting (WRF) model is a numerical weather prediction and atmospheric simulation system. It has been designed for both research and operational applications. WRF code can be run in different computing environments ranging from laptops to supercomputers. Purdue Lin scheme is a relatively sophisticated microphysics scheme in WRF. The scheme includes six classes of hydro meteors: water vapor, cloud water, raid, cloud ice, snow and graupel. In this paper, we accelerate the Purdue Lin scheme on the multi-core NVIDIA Graphics Processing Units (GPUs). Lately, GPUs have evolved into highly parallel, multi-threaded, many-core processors possessing tremendous computational speed and a high memory bandwidth. We discuss how our GPU implementation exploits the massive parallelism, resulting in a highly efficient acceleration of the Purdue Lin scheme. We utilize a low-cost personal supercomputer with 512 CUDA cores on a GTX590 GPU. We achieve an overall speedup of 156× in case of 1 GPU as compared to the single-threaded CPU version. Since Purdue Lin microphysics scheme is only an intermediate module of the entire WRF model, host-device I/O should not happen, i.e. its input data is already available in the GPU global memory from previous modules and its output data should reside in the GPU global memory for later usage by other modules. The speedup without host-device data transfer time is 692×.

The weather research and Forecasting (WRF) model in an atmospheric simulation system, which is designed for both operational and research use. This common tool aspect promotes closer ties between research and operational communities. It contains a lot a different physics and dynamics options reflecting the experience and input of the broad scientific community. The WRF physics categories and microphysics, cumulus parametrization, planetary boundary layer, land-surface model and radiation. Explicitly resolved water vapor, cloud and precipitation processes are included in microphysics. Several bulk water microphysics schemes are available within the Weather Research and Forecasting (WRF) model, with different numbers of simulated hydrometeor classes and methods for estimating their size fall speeds, distributions and densities. Stony-Brook University (SBU-YLIN) microphysics scheme is a 5-class scheme with riming intensity predicted to account for mixed-phase processes. In this paper, we develop an efficient graphics processing unit (GPU) based SBU-YLIN scheme. WRF computation domain is 3D grid layed over the earth. SBU-YLIN performs the same computation for each spatial position in the whole domain. This repletion of the same computation on different data sets allows using GPU's Single Instruction Multiple Dataset (SIMD) architecture. The GPU-based SBUYLIN scheme will be compared to a CPU-based single-threaded counterpart. The implementation achieves 213x speedup with I/O compared to a Fortran implementation running on a CPU. Without I/O the speedup is 896x.

Layer-mean virtual temperatures retrieved from satellite measurements are more accurate than retrievals at specific pressure not only because an averaging process is involved, but also because of advantages in the retrieval process. In this note, a “retrieval efficiency” is derived to express this advantage over simple averaging as a function of layer thickness. The efficiency is examined for two common cases of retrieval initial guess: a statistical sample mean and a forecast profile obtained from a numerical prediction model. The advantage of the layer-mean retrieval clearly is demonstrated in both cases.

Contemporary and future ultraspectral sounders represent a significant technical advancement for environmental and meteorological prediction and monitoring. Given their large volume of spectral observations, the use of robust data compression techniques will be beneficial to data transmission and storage. In this paper, we propose a novel Adaptive Vector Quantization (VQ)-based Linear Prediction (AVQLP) method for ultraspectral data compression. The method is compared with several state-of-the-art methods such as CALIC, JPEG-LS and JPEG2000. The compression experiments show that our AVQLP method is the first to surpass the 4 to 1 lossless compression barrier for a selected set of AIRS ultraspectral sounder test data.

Previous study shows 3D Wavelet Transform with Reversible Variable-Length Coding (3DWT-RVLC) has much better error resilience than JPEG2000 Part 2 on 1-bit random error remaining after channel decoding. Errors in satellite channels might have burst characteristics. Low-density parity-check (LDPC) codes are known to have excellent error correction capability near the Shannon limit performance. In this study, we investigate the burst error correction performance of LDPC codes via the new Digital Video Broadcasting - Second Generation (DVB-S2) standard for ultraspectral sounder data compressed by 3DWT-RVLC. We also study the error contamination after 3DWT-RVLC source decoding. Statistics show that 3DWT-RVLC produces significantly fewer erroneous pixels than JPEG2000 Part 2 for ultraspectral sounder data compression.

Most source coding techniques generate bitstream where different regions have unequal influences on data reconstruction. An uncorrected error in a more influential region can cause more error propagation in the reconstructed data. Given a limited bandwidth, unequal error protection (UEP) via channel coding with different code rates for different regions of bitstream may yield much less error contamination than equal error protection (EEP). We propose an optimal UEP scheme that minimizes error contamination after channel and source decoding. We use JPEG2000 for source coding and turbo product code (TPC) for channel coding as an example to demonstrate this technique with ultraspectral sounder data. Wavelet compression yields unequal significance in different wavelet resolutions. In the proposed UEP scheme, the statistics of erroneous pixels after TPC and JPEG2000 decoding are used to determine the optimal channel code rates for each wavelet resolution. The proposed UEP scheme significantly reduces the number of pixel errors when compared to its EEP counterpart. In practice, with a predefined set of implementation parameters (available channel codes, desired code rate, noise level, etc.), the optimal code rate allocation for UEP needs to be determined only once and can be done offline.

Multiwavelets are theoretically expected to perform better than traditional wavelets for image compression. Standard and new decomposition methods for multiwavelet compression have been studied in (1), where floating-point multiwavelet transforms are used for lossy image compression. In this paper we investigate both decomposition methods with a reversible integer multiwavelet transform for lossless compression of ultraspectral sounder data. This study shows that the multiwavelet compression method outperforms other well-known methods such as CALIC, SPIHT, and CCSDS IDC for lossless compression on ultraspectral data. Furthermore, the standard multiwavelet decomposition method yields better compression than the new one. We also show that the Bias- Adjusted Reordering (BAR) scheme for spectral data preprocessing yields a substantial improvement on multiwavelet compression with the average compression ratio increasing from 2.27 to 2.61.

Reversible variable-length codes (RVLCs) allow instantaneous decoding in both directions, which affords better detection of bit errors due to synchronization losses over a noisy channel. Earlier, we developed 3D wavelet reversible variable- length coding (3DWT-RVLC) for lossless compression of ultraspectral sounder data, which has significantly better error resilience than JPEG2000 Part 2 at only a small reduction in compression gain (1). To explore the feasibility of 3DWT-RVLC for real-time satellite onboard processing, we implement a memory-limited DSP version of 3DWT-RVLC. Experimental results for the 10 AIRS ultraspectral test granules show that the DSP-based 3DWT-RVLC yields an average compression ratio of 2.36, comparable to 2.51 of the original 3DWT-RVLC from our previous work.

The ultraspectral sounder data features strong correlations in disjoint spectral regions due to the same type of absorbing gases. This paper compares the compression performance of two robust data preprocessing schemes, namely Bias-Adjusted reordering (BAR) and Minimum Spanning Tree (MST) reordering, in the context of entropy coding. Both schemes can take advantage of the strong correlations for achieving higher compression gains. The compression methods consist of the BAR or MST preprocessing schemes followed by linear prediction with context-free or context-based arithmetic coding (AC). Compression experiments on the NASA AIRS ultraspectral sounder data set show that MST without bias-adjustment produces lower compression ratios than BAR and bias-adjusted MST for both context-free and context-based AC. Biasadjusted MST outperforms BAR for context-free arithmetic coding, whereas BAR outperforms MST for context-based arithmetic coding. BAR with context-based AC yields the highest average compression ratios in comparison to MST with context-free or context-based AC.

Low density parity check (LDPC) codes have the ability to transmit data at rates approaching channel capacity with very low error probability. Our study of LDPC codes on JPEG2000-compressed ultraspectral sounder data shows that for an LDPC encoder designed with certain input parameters, the LDPC decoder is able to correct all single errors and two-bit burst errors. Furthermore, our analysis on a massive variety of 3- and 4-bit burst error scenarios indicates that the locations of LDPC-uncorrected errors for 3- and 4-bit burst errors appear to be fixed, regardless of the content of the input data. We propose a look-up table based error correction method to correct errors surviving from the LDPC decoder for 3- and 4-bit burst errors.

PCA is a powerful tool for AIRS data compression. Eigenvector from a dependent dataset works best for data compression; those from independent datasets can be used to diagnose climate abnormal signals or high-noise in observations.

In this paper, we consider an application of the Empirical Mode Decomposition (EMD) introduced by Norden E. Huang in 1996 to the compression of 3D hyperspectral sounding data. The EMD is a new data analysis method which is based on expansion of the data in terms of Intrinsic Mode Functions (IMF). These IMFs are based on and derived from the data set. Since EMD adaptively represent the signal as a sum of "well behaved" amplitude/frequency modulated components, we found it very well suited for the whitening part of the compression scheme.

Future satellites will have the higher spectral resolutions required to provide an adequate level of vertical resolution in the soundings they produce, but the high resolution is accompanied by a large number of channels. In order for the data to be used by numerical prediction centers, the number must be reduced while maintaining the information content. This paper presents an approach for reducing the number of channels by averaging similar channels to produce “super channels” with reduced noise. To be useful, the “super channels” must be accompanied by an fast and accurate approximation of the Plank function. Up till now, this has been a limitation in the application of the approach. This paper presents a solution for adjusting the calculation at a single frequency to match the value for an average over a spectral interval. The approach will be applied to AIRS channels and the results will be presented at the meeting.

Pattern recognition procedures have been developed in order to improve the first-guess fields for satellite temperature retrievals. The first procedure is used to select one or more historical radiosonde temperature profiles as analog estimates of ambient thermal structure. The second procedure is used to organize a priori data into shape-coherent pattern libraries using structural information inherent in the data itself. On the basis of independent tests of about 800 temperature retrievals, it was found that: (1) the pattern recognition techniques reduced first-guess profile errors by nearly 50 percent in comparison with traditional partitioning schemes; and (2) with regression and physical-iterative retrieval algorithms, however, the effect of pattern recognition on temperature retrieval error was insignificant. Analysis of individual retrieval errors showed that poor retrievals may outweigh the potential benefits of both pattern recognition techniques.