Decentralized and coordinate-free computation of critical points and surface networks in a discretized scalar field     

Myeong-Hun Jeong  Matt Duckham  Allison Kealy  Harvey J. Miller  Andrej Peisker 《International journal of geographical information science》2014,28(1):1-21

This article provides a decentralized and coordinate-free algorithm, called decentralized gradient field (DGraF), to identify critical points (peaks, pits, and passes) and the topological structure of the surface network connecting those critical points. Algorithms that can operate in the network without centralized control and without coordinates are important in emerging resource-constrained spatial computing environments, in particular geosensor networks. Our approach accounts for the discrepancies between finite granularity sensor data and the underlying continuous field, ignored by previous work. Empirical evaluation shows that our DGraF algorithm can improve the accuracy of critical points identification when compared with the current state-of-the-art decentralized algorithm and matches the accuracy of a centralized algorithm for peaks and pits. The DGraF algorithm is efficient, requiring O(n) overall communication complexity, where n is the number of nodes in the geosensor network. Further, empirical investigations of our algorithm across a range of simulations demonstrate improved load balance of DGraF when compared with an existing decentralized algorithm. Our investigation highlights a number of important issues for future research on the detection of holes and the monitoring of dynamic events in a field.  相似文献   

Assessing drought threats to agricultural water supplies under climate change by combining the SWAT and MODSIM models for the Geum River basin,South Korea     

So Ra Ahn  Jae Hak Jeong 《水文科学杂志》2013,58(15):2740-2753

ABSTRACT

The impacts of future climate change on the agricultural water supply capacities of irrigation facilities in the Geum River basin (9645.5 km²) of South Korea were investigated using an integrated modeling framework that included a water balance network model (MODSIM) and a watershed-scale hydrologic model (Soil and Water Assessment Tool, SWAT). The discharges and baseflows from upland drainage areas were estimated using SWAT, and the predicted flow was used to feed agricultural reservoirs and multipurpose dams in subwatersheds. Using a split sampling method, we calibrated the daily streamflows and dam inflows at three locations using data from 6 years, including 3 years of calibration data (2005–2007) followed by 3 years of validation data (2008–2010). In the MODSIM model, the entire basin was divided into 14 subwatersheds in which various agricultural irrigation facilities such as agricultural reservoirs, pumping stations, diversions, culverts and groundwater wells were defined as a network of hydraulic structures within each subwatershed. These hydraulic networks between subwatersheds were inter-connected to allow watershed-scale analysis and were further connected to municipal and industrial water supplies under various hydrologic conditions. Projected climate data from the HadGEM3-RA RCP 4.5 and 8.5 scenarios for the period of 2006–2099 were imported to SWAT to calculate the water yield, and the output was transferred to MODSIM in the form of time-series boundary conditions. The maximum shortage rate of agricultural water was estimated as 38.2% for the 2040s and 2080s under the RCP 4.5 scenario but was lower under the RCP 8.5 scenario (21.3% in the 2040s and 22.1% in the 2080s). Under the RCP 4.5 scenario, the projected shortage rate was higher than that during the measured baseline period (1982–2011) of 25.6% and the RCP historical period (1982–2005) of 30.1%. The future elevated drought levels are primarily attributed to the increasingly concentrated rainfall distribution throughout the year under a monsoonal climate, as projected by the IPCC climate scenarios.