Along with the gradually accelerated urbanization process, simulating and predicting the future pattern of the city is of great importance to the prediction and prevention of some environmental, economic and urban issues. Previous studies have generally integrated traditional machine learning with cellular automaton (CA) models to simulate urban development. Nevertheless, difficulties still exist in the process of obtaining more accurate results with CA models; such difficulties are mainly due to the insufficient consideration of neighborhood effects during urban transition rule mining. In this paper, we used an effective deep learning method, named convolution neural network for united mining (UMCNN), to solve the problem. UMCNN has substantial potential to get neighborhood information from its receptive field. Thus, a novel CA model coupled with UMCNN and Markov chain was designed to improve the performance of simulating urban expansion processes. Choosing the Pearl River Delta of China as the study area, we excavate the driving factors and the transformational relations revealed by the urban land-use patterns in 2000, 2005 and 2010 and further simulate the urban expansion status in 2020 and 2030. Additionally, three traditional machine-learning-based CA models (LR, ANN and RFA) are built to attest the practicality of the proposed model. In the comparison, the proposed method reaches the highest simulation accuracy and landscape index similarity. The predicted urban expansion results reveal that the economy will continue to be the primary factor in the study area from 2010 to 2030. The proposed model can serve as guidance in urban planning and government decision-making. 相似文献
After the bursting of Huiten Nor in Hoh Xil Region in September, 2011, the topic on whether the water overflowed from the Salt Lake would enter into the Chumaer River and become the northernmost source of the Yangtze River has aroused wide concern from public and academic field. Based on Landsat TM/ETM+/OLI remote sensing images during 2010–2015, SRTM 1 arc-second data, Google Earth elevation data and the observation data from the Wudaoliang meteorological station, the study initially analyzed the variations of the Salt Lake and its overflowing condition and probability. The results showed that the area of the Salt Lake expanded sharply from October 2011 to April 2013, and then it stepped into a stable expansion period. On October 27, 2015, the area of the Salt Lake had arrived at 151.38 km2, which was about 3.35 times the area of the lake on March 3, 2010. The Salt Lake will overflow when its area reaches the range from 218.90 km2 to 220.63 km2. Due to the differences between SRTM DEM and Google Earth elevation data, the water level of the Salt Lake simulated would be 12 m or 9.6 m higher than the current level when the lake overflowed, and its reservoir capacity would increase by 23.71 km3 or 17.27 km3, respectively. Meanwhile, the overflowed water of the Salt Lake would run into the Qingshui River basin from its eastern part. Although the Salt Lake does not overflow in the coming decade, with watershed expansion of the Salt Lake and the projected precipitation increase in Hoh Xil region, the probability of water overflow from the Salt Lake and becoming a tributary of the Yangtze River will exist in the long term. 相似文献
The distinctive estuary hydrodynamics and nutrient input make the estuary ecosystem play a key role in lake ecosystems. The Nanfei River and Zhaohe River are two main inlets of Chaohu Lake, Anhui, East China. We selected estuaries of the two rivers as representative areas to study temporal and spatial changes of bacterial communities. In August (summer) and November (autumn) 2016 and February (winter) and May (spring) 2017, 16 water and sediment samples were collected from the estuaries. Physicochemical characteristics indicate significant differences in the nutritional status and eutrophication index of the estuaries due mainly to organic input. Examination of the number of operational taxonomic units, the diversity index, the community composition, and redundancy analysis revealed the following. First, the existence of varying degrees of seasonal differences in the distribution of almost all bacteria. In addition, the species diversity in the sediment samples was higher than that in the water samples, and the dominant species differed also among these samples. Second, a large number of unknown genera were detected, especially in the sediment samples, such as unclassified Xanthomonadales incertae sedis, unclassified Anaerolineaceae, and unclassified Alcaligenaceae. Last, TP, TN, and TOC were the main influential factors that affected the bacterial community structure. 相似文献
The eastern edge of the Qinghai-Tibetan plateau developed an integrated series of late Cenozoic lacustrine, loess, red and
moraines deposits. Various genetic sediments recorded rich information of Quaternary palaeoenvironment changes. Xigeda Pliocene
lacustrine deposits, formed during 4.2 Ma B.P.–2.6 Ma B.P., experienced nine periodic warm-cold stages. Eolian deposition
in western Sichuan began at 1.15 Ma B.P., and the loess-soil sequences successively record fourteen palaeomonsoon change cycles.
Red clay in the Chengdu plain record five stages of paleoclimatic change stages since 1.13 Ma B.P.. There was an old glacial
period of 4.3 Ma B.P. in the eastern Qinghai-Tibetan plateau. During the Quaternary, there were five extreme paleoclimatic
events corresponding to five glaciations.
__________
Translated from Geological Bulletin of China, 2007, 26(12): 1620–1626 [译自: 地质通报] 相似文献
Accurate simulation of rice yield is very important and vital for agriculture and food security. This study has analyzed the applicability of the RS-P-YEC (the remote-sensing-photosynthesis-yield estimation for crops) model for the rice yield simulation of the Middle and Lower Reaches of Yangtze River (MLRYR) in China. The simulated rice yield was compared with the actual statistical dataset, so as to obtain the accuracy of the model results. The results showed that the correlation coefficients (R) between simulated rice yield and statistical data is 0.708 (P < 0.01), the average relative errors were 9, 6.5, 7.2 %, and the root mean square errors were 777.5, 606.4, 693.4 kg/ha in 2007, 2008 and 2009, respectively. It indicated that the RS-P-YEC model can be used to estimate rice yield in the MLRYR region of China. 相似文献
Soil moisture estimation from satellite earth observation has emerged effectively advantageous due to the high temporal resolution, spatial resolution, coverage, and processing convenience it affords. In this paper, we present a study carried out to estimate soil moisture level at every location within Enugu State Nigeria from satellite earth observation. Comparative analysis of multiple indices for soil moisture estimation was carried out with a view to evaluating the robustness, correlation, appropriateness and accuracy of the indices in estimating the spatial distribution of soil moisture level in Enugu State. Results were correlated and validated with In-Situ soil moisture observations from multi-sample points. To achieve this, the Topographic Wetness Index (TWI), based on digital elevation data, the Temperature Vegetation Dryness Index (TVDI) and an improved TVDI (iTVDI) incorporating air temperature and a Digital Elevation Model (DEM) were calculated from ASTER global DEM and Landsat images. Possible dependencies of the indices on land cover type, topography, and precipitation were explored. In-Situ soil moisture data were used to validate the derived indices. The results showed that there was a positive significant relationship between iTVDI versus TVDI (R = 0.53, P value < 0.05), while in iTVDI versus TWI (R = 0.00, P value > 0.05) and TVDI versus TWI (R = ?0.01, P value > 0.05) no significant relationship existed. There was a strong relationship between iTVDI and topography, land cover type, and precipitation than other indices (TVDI, TWI). In situ measured soil moisture values showed negative significant relationship with TVDI (R = ?0.52, P value < 0.05) and iTVDI (R = ?0.63, P value < 0.05) but not with TWI (R = ?0.10, P value > 0.05). The iTVDI outperformed the other two index; having a stronger relationship with topography, precipitation, land cover classes and soil moisture. It concludes that although iTVDI outperformed other indices (TVDI, TWI) in soil moisture estimation, the decision of which index to apply is dependent on available data, the intent of usage and spatial scale. 相似文献
Global increases in duration and prevalence of droughts require detailed drought characterization at various spatial and temporal scales. In this study, drought severity in Xinjiang, China was investigated between 1961 and 2012. Using meteorological data from 55 weather stations, the UNEP (1993) index (IA), Erinç’s aridity index (Im), and Sahin’s aridity index (Ish) were calculated at the monthly and annual timescales and compared to the Penman-Monteith based standard precipitation evapotranspiration index (SPEIPM). Drought spatiotemporal variability was analyzed for north (NX), south (SX), and entire Xinjiang (EX). Im could not be calculated at 51 stations in winter as Tmax was below 0. At the monthly timescale, IA, Im, and Ish correlated poorly to SPEIPM because of seasonality and temporal variability, but annual IA, Im, and Ish correlated well with SPEIPM. Annual IA, Im, and Ish showed strong spatial variability. The 15 extreme droughts denoted by monthly SPEIPM occurred in NX but out of phase in SX. Annual precipitation, maximum temperature, and relative and specific humidity increased, while air pressure and potential evapotranspiration decreased over 1961–2012. The resulting increases in the four drought indices indicated that drought severity in Xinjiang decreased, because the local climate became warmer and wetter.