Chaolin Wu

Soil moisture retrieval from MODIS data in Northern China Plain using thermal inertia model

Soil moisture plays an important role in surface energy balances, regional runoff, potential drought and crop yield. Early detection of potential drought or flood is important for the local government and people to take actions to protect their crop. Traditionally measurement of soil moisture is a time‐consuming job and only limited samples could be collected. Many problems would be results from extending those point measurements to 2D space, especially for a regional area with heterogeneous soil characteristics. The emergency of remote‐sensing technology makes it possible to rapidly monitor soil moisture on a regional scale. Thermal inertia represents the ability of a material to conduct and store heat, and in the context of planetary science, it is a measure of the subsurfaces ability to store heat during the day and reradiate it during the night. One major application of thermal inertia is to monitor soil moisture. In this paper, a thermal inertia model was developed to be suitable in situations whether or not the satellite overpass time coincides with the local maximum and minimum temperature time. Besides, the sensibilities of thermal inertia with surface albedo and the surface temperature difference were discussed. It shows that the surface temperature difference has more effects on the thermal inertia than the surface albedo. When the temperature difference is less than 10 Kelvin degrees, 1 Kelvin degree error of temperature difference will lead to a big fluctuation of thermal inertia. When the temperature difference is more than 10 Kelvin degrees, 1 Kelvin degree error of temperature difference will cause a small change of thermal inertia. The temperature difference should be larger than 10 Kelvin degrees when the thermal inertia model is selected to derive soil moisture or other applications. Based on this thermal inertia model, the soil moisture map was obtained for North China Plain. It shows that the averaged difference between the soil moisture values derived from MODIS data and in situ measured soil moisture data is 4.32%. This model is promising for monitoring soil moisture on a large regional scale.

Preliminary study of Grid computing for remotely sensed information

Observing the world‐wide concentration and distribution of ozone is important for monitoring the evolution of the ozone layer, to derive the amount of UV, to provide ozone and UV forecasts and to improve weather forecasting. Monitoring ozone is the primary function of the Global Ozone Monitoring Experiment. Each day, space missions download, from space to ground, many raw images that are stored in ground stations located all over the world. How to process this data resource in real time — or almost real time — and effectively share spatial information among the remote sensing community is a pressing task. Grid computing can provide access to a globally distributed computing environment via authentication, authorization, negotiation and security. It can create a computational environment handling many PetaBytes of geographically distributed data, tens of thousands of heterogeneous computing resources and thousands of simultaneous users from many research institutions. It can provide a powerful tool for sharing both remote sensing data and processing middleware. This paper introduces the concept of grid computing, followed by its applications for atmospheric ozone retrieval. The special remote sensing data analysis note for the Spatial Information Grid (SIG) is addressed in detail. A series of remotely sensed image processing middleware is shown. Experience shows that near‐real‐time products, such as maps of ozone, from the processing and analysis of remotely sensed data are possible.

A remote sensing application workflow and its implementation in remote sensing service grid node

In this article we describe a remote sensing application workflow in building a Remote Sensing Information Analysis and Service Grid Node at Institute of Remote Sensing Applications based on the Condor platform. The goal of the Node is to make good use of physically distributed resources in the field of remote sensing science such as data, models and algorithms, and computing resource left unused on Internet. Implementing it we use workflow technology to manage the node, control resources, and make traditional algorithms as a Grid service. We use web service technology to communicate with Spatial Information Grid (SIG) and other Grid systems. We use JSP technology to provide an independent portal. Finally, the current status of this ongoing work is described.

Java-Based grid service spread and implementation in remote sensing applications

Remote sensing applications often concern very large volumes of spatio-temporal data, the emerging Grid computing technologies bring an effective solution to this problem. The Open Grid Services Architecture (OGSA) treats Grid as the aggregate of Grid service, which is extension of Web Service. It defines standard mechanisms for creating, naming, and discovering transient Grid service instances; provides location transparency and multiple protocol bindings for service instances; and supports integration with underlying native platform facilities. It is not effective used in data-intensive computing such as remote sensing applications because its foundation, Web Service, is not efficient in scientific computing. How to increase the efficiency of the grid services for a scientific computing? This paper proposes a mechanism Grid service spread (GSS), which dynamically replant a Grid service from a Grid node to the others. We have more computers to provide the same function, so less time can be spent completing a problem than original Grid system. This paper also provides the solution how to adept the service duplicate for the destination node’s Grid environment; how each service duplicate communicates with each other; how to manage the lifecycle of services spread etc. The efficiency of this solution through a remote sensing application of NDVI computing is demonstrated. It shows that this method is more efficient for processing huge amount of remotely sensed data.

Remote sensing information processing grid node with loose-coupling parallel structure

To use traditional algorithms and software packages on Grid system, traditional algorithms and software packages, in general, have to be modified. In this paper we focus on standards and methodologies for Grid platform within the context of the Remote Sensing Data Processing Grid Node (RSDPGN) that implements a loose-coupling parallel structure for orchestrating traditional remote sensing algorithms and software packages on the Condor platform. We have implemented 17 remote sensing applications in one system using Web service and workflow technology without any change to traditional codes. Some core algorithm codes are come from a remote sensing software package which we has neither resource codes nor APIs. Others come from the program codes accumulated by our group. The design and prototype implementation of RSDPGN are presented. The advantage and shortage of loose-coupling structure is analysed. Through a case study of land surface temperature calcu-lation from MODIS data, we demonstrate the way to modify software packages in details. Moreover we discuss the problems and solutions based on our experience such as system architecture, the kinds of functional modules, fast data transfer, and state monitoring.

Study on grid-based special remotely sensed data processing node in grid GIS

Grid Geospatial Information Service (Grid GIS) system aims to study and develop grid-based uniform spatial information access and analysis system. Data resources of Grid GIS include not only original and traditional GIS data but remotely sensed data. Everyday, space missions involve the download, from space to ground, of huge amount of raw images that are stored in the ground stations geographically distributed. It is a practical pressing task to process these data resource in real time or almost real time and to effectively share spatial information among remote sensing community. Grid technology can provide access to a global distributed computing environment via authentication, authorization, negotiation and security tools. This paper discusses the key technologies of Grid-based special remotely sensed data processing node. First, the concept of Grid-based special remotely sensed data processing node is introduced. Following is the architecture and functions of this node. Based on this architecture, the tasks scheduling algorithm is presented. Finally we introduce the computing resource meta-module information registry and renewal mechanism of the Grid-based special remotely sensed data processing node for Grid GIS.

A study of grid workflow dynamic customization for remote sensing quantitative retrieval

A grid workflow is a type of high-level grid middleware to support modeling, redesign and execution of large-scale intricate scientific and business processes in many complex e-science applications. In this paper, we make improvement on existing grid workflow to meet the application request of remote sensing in two ways: expand the interface of models and design map of remote sensing tasks on grid platform. A case of remote sensing quantitative retrieval grid workflow is shown in section 3.

Prior information supported aerosol optical depth retrieval using FY2D data

The algorithm is based on the assumption that TOA reflectance increase with the aerosol load as well as the surface reflectance at same time gradually changes on different days within 14 days. Then the surface reflectance is derived from FengYun-2D (FY2D) measurements every 1 hour as the second darkest of reflectance for each time of day to minimize the effect of geometry change and cloud. The “true surface reflectance” of each time was calculated from the composite reflectance and their weighs. The weigh of each time, contribution of the surface and aerosol background were determined using the prior information, and both of them were various in different time. The AOD retrieval based on a Look-Up Table (LUT) using composite background (CB) method and improved composite background (ICB) algorithm were compared with AERONET sites, it was found that the ICB provides larger coverage and higher accuracy AOD product compared with CB.

Grid service implementation of aerosol optical thickness retrieval over land from MODIS

To derive the actual land surface information quantitatively, the atmospheric effects should be correctly removed. Atmospheric effects dependent on aerosol particles, clouds and other atmosphere conditions. Aerosol parameters can be retrieved from the remotely sensed data. The retrieved aerosol characters can also be applied to environmental monitoring. To retrieval the aerosol optical thickness over land, many methods have been developed. The most popular one is the dark dense vegetation method. But it is confined to vegetation fields. The SYNTAM method can be used to retrieval aerosol optical thickness over land from MODIS data, no matter whether the land is dark or bright. In this paper, the SYNTAM method is applied to MODIS data for the retrieval of aerosol optical thickness over China. The retrieval process is complicated. And the EMS memory required is too large for a personal computing to run successfully. To solve this problem, the Grid environment is used. Our experiments were performed on the High-Throughput Spatial Information Processing Prototype System based on Grid platform in Institute of Remote Sensing Applications, Chinese Academy of Sciences. The aerosol optical thickness retrieval process is described in this paper. And the detail data query, data pre-processing, job monitoring and post-processing is discussed. Moreover, test results are also reported in this paper.

Study of task managing strategy in remote sensing information analysis and service grid node

A huge amount of remotely sensed data is acquired daily and many algorithms for processing the data are complex. Grid computing, which is different from conventional distributed computing, can harness the power of many computers in a network to solve problems. Grid computing has become an exciting way of solving the large-scale or complex problems. To apply the grid computing to remote sensing information, a remote sensing information grid analysis and service node (RSIGN) has been established in Telegeoprocessing Group, Institute of Remote Sensing Applications, Chinese Academy of Sciences. The management and distribution of the tasks are key problems in RSIGN node. How to distribute and manage the tasks has significant influence on the efficiency of the whole grid system. In this paper, we will discuss two main strategies for RSIGN node: geometric parallel and algorithm parallel. Different problems in the grid node need different strategies. On one hand, the data to be processed and analyzed can be divided into sub-datasets and processed on different computers. On the other hand, some algorithms can be rewritten to fit the parallel rules. In the latter case, different steps of processing the whole dataset can be executed on different computers. Through the experience of applied the both strategies on RSIGN node, we compared and evaluated the two strategies, and figured out the merits and disadvantages of two different strategies when applied on remotely sensed information analysis. At the end of the paper, we discuss how to choose the best task managing strategy for different problems in remote sensing information processing and analysis.