Information Gathering on Rice Planting Area Using GF-1/WFV EVI Time Series Technology
-
摘要: 随着国产高分系列卫星的发射，国产高分辨率遥感影像在作物种植面积提取方面逐渐得到了应用，但目前仅有少量基于高分一号（GF-1）卫星影像归一化植被指数（Normalized Difference Vegetation Index，NDVI）进行作物分类或遥感估产的研究。本文针对NDVI指数在南方多植被覆盖区域易饱和及不敏感的缺陷，利用在NDVI基础上改进的增强型植被指数（Enhanced Vegetation Index，EVI），首次提出基于GF-1卫星16 m宽覆盖影像（Wide Field of View，WFV）EVI时间序列的水稻种植面积提取方法。选取四川省乐至县为研究区域，获取覆盖整个水稻生长周期的GF-1/WFV影像数据，构建EVI时间序列，并分析水稻不同生长期的EVI曲线特征。利用时间序列谐波分析法（Harmonic Analysis of Time Series，HANTS）对EVI时间序列进行平滑处理，尽可能减少噪声影响，使EVI时间序列能够更好地反应水稻及其他植被或非植被随时间的变化规律；根据水稻和其他植被及非植被的EVI曲线特征差异构建水稻种植面积提取决策树模型，对水稻种植面积进行准确提取。通过与同期地理国情监测成果的对比，本研究方法提取的水稻种植面积和精度都较高，表明该方法对于提取水稻种植面积效果良好。研究表明，相对于以往用中低分辨率卫星影像进行作物种植结构提取，GF1/WFV影像在南方较破碎的水稻种植面积提取方面应用效果良好，GF-1卫星影像在农业遥感领域具备很大的应用潜力。Abstract: With the successful launching of high-resolution satellites of the China High-Resolution Earth Observation System, the remote sensing images obtained will be increasingly applied for information gathering on crop planting. At present, limited studies were conducted on crop classification and yield estimation using the GF-1 satellite images based on the Normalized Difference Vegetation Index(NDVI). The NDVI transmissions tended to be over-crowded and low in resolution for the southern part of China due to the high vegetation growth in the areas.The proposed method utilized the improved Enhanced Vegetation Index(EVI) time series of GF-1 Wide Field View (WFV) images to overcome the deficiency. Lezhi county in Sichuan province was chosen for our testing.The GF-1/WFV images encompassing the entire rice growth period were acquired to construct the EVI time series.The characteristic EVI curves were obtained for the rice crops covering different growth stages. The Harmonic Analysis of Time Series(HANTS) method was adopted to smooth the EVI time series and maximally reduce the noise to enable reliable reflection of the dynamic changing patterns of rice and other crops or non-crop objects. Thereby, with the aid of the decision tree model, the rice planting area and other relevant information could be extracted. Comparing the results with what was obtained by the geographical conditions monitoring during a same time frame, the current method was considered accurate and precise. The GF-1/WFV images were particularly superior to the moderate or low resolution satellite images for gathering information on rice planting area in the regions where rice fields are scattered widely. The GF-1 technology was believed to represent a significant potential for applications in the field of agricultural remote sensing.
-
Key words:
- GF1 /
- EVI /
- time series /
- rice /
- area /
- geographical conditions monitoring
-
图 3 水稻种植面积提取决策树模型
Figure 3. Decision tree model for information extraction on rice planting area
图 6 乐至县水田分布
注：据2016年地理国情监测。
Figure 6. Spatial distribution of rice planting areas in Lezhi county
表 1 GF-1/WFV影像信息
Table 1. Information of GF-1/WFV images
序号 传感器类型 中心经、纬度 获取时间/(年-月-日) 1 WFV4 104.3°E, 30.1°N 2016-03-18 2 WFV1 104.9°E, 29.6°N 2016-05-09 3 WFV4 104.1°E, 30.2°N 2016-06-16 4 WFV2 105.7°E, 31.0°N 2016-08-24 5 WFV2 105.3°E, 29.3°N 2016-08-24 6 WFV4 105.1°E, 30.2°N 2016-09-10 7 WFV1 105.4°E, 29.6°N 2016-11-26 
下载: 导出CSV
表 2 水稻提取精度评价
Table 2. Precision of information extraction on rice planting area
类别 水稻/% 非水稻/% 总体精度/% Kappa系数 水稻 96.73 3.77 96.52 0.93 非水稻 3.27 96.23 制图精度 96.73 96.23 用户精度 97.28 95.48
下载: 导出CSV
-
[1] 陈仲新, 任建强, 唐华俊, 等.农业遥感研究应用进展与展望[J].遥感学报, 2016(5):748-767. http://www.cnki.com.cn/Article/CJFDTOTAL-WHCH201405001.htm [2] 吴炳方.全国农情监测与估产的运行化遥感方法[J].地理学报, 2000(1):25-35. doi: 10.11821/xb200001003 [3] 杨晓光, 刘志娟, 陈阜.全球气候变暖对中国种植制度可能影响Ⅰ.气候变暖对中国种植制度北界和粮食产量可能影响的分析[J].中国农业科学, 2010, 43(2):329-336. http://www.cnki.com.cn/Article/CJFDTOTAL-ZNYK201002015.htm [4] BENEDETTI R, ROSSINI P. On the use of NDVI profiles as a tool for agricultural statistics:The case study of wheat yield estimate and forecast in Emilia Romagna[J]. Remote sensing of environment, 1993, 45(3):311-326. doi: 10.1016/0034-4257(93)90113-C [5] MACDONALD R B, HALL F G. Global Crop Forecasting[J]. Science, 1980, 208(4445):670-679. doi: 10.1126/science.208.4445.670 [6] CARLSON T N, RIPLEY D A. On the relation between NDVI, fractional vegetation cover, and leaf area index[J]. Remote sensing of environment, 1997, 62(3):241-252. doi: 10.1016/S0034-4257(97)00104-1 [7] 王延颐, MALINGREAU J P.应用NOAA-AVHRR对江苏省作物进行监测的可行性研究[J].环境遥感, 1990, 5(3):221-227. http://www.oalib.com/paper/4184007 [8] 肖乾广.用NOAA气象卫星AVHRR的定量资料计算冬小麦种植面积的两种方法[J].环境遥感, 1989, 4(3):191-196. http://www.cnki.com.cn/Article/CJFDTotal-ZGNY199204005.htm [9] 彭代亮. 基于统计与MODIS数据的水稻遥感估产方法研究[D]. 杭州: 浙江大学, 2009. http://www.wanfangdata.com.cn/details/detail.do?_type=degree&id=Y1647020 [10] 张军龙. 基于MODIS_EVI重庆水稻估产[D]. 重庆: 重庆师范大学, 2014. http://cdmd.cnki.com.cn/Article/CDMD-10637-1014357388.htm [11] 王长耀.林文鹏.基于MODIS EVI的冬小麦产量遥感预测研究[J].农业工程学报, 2005, 21(10):90-94. doi: 10.3321/j.issn:1002-6819.2005.10.021 [12] 丁春雨, 马冠南, 张晓娇, 等.利用高分一号影像提取水稻种植面积方法研究——以长春榆树市为例[J].农业与技术, 2016, 36(7):1-4, 19. http://d.old.wanfangdata.com.cn/Periodical/nyyjs201607002 [13] 贾玉秋, 李冰, 程永政, 等.基于GF-1与Landsat-8多光谱遥感影像的玉米LAI反演比较[J].农业工程学报, 2015, 31(9):173-179. doi: 10.11975/j.issn.1002-6819.2015.09.027 [14] 杨闫君, 占玉林, 田庆久, 等.基于GF-1/WFVNDVI时间序列数据的作物分类[J].农业工程学报, 2015, 31(24):155-161. doi: 10.11975/j.issn.1002-6819.2015.24.024 [15] ROUSE J W, HAAS R H, SCHELL J A, et al. Monitoring vegetation systems in the Great Plains with ERTS[C]. //NASA. Goddard Space Flight Center Third Earth Resources Technology Satellite (ERTS) symposium, 1974(1), Sect. A: 309-317. [16] 王正兴, 刘闯, ALFREDO H.植被指数研究进展:从AVHRR-NDVI到MODIS-EVI[J].生态学报, 2003, 23(5):979-987. http://www.cnki.com.cn/Article/CJFDTOTAL-STXB200305019.htm [17] HUETE A R, LIU H Q, BATCHILY K, et al. A comparison of vegetation indices over a global set of TM images for EOS-MODIS[J]. Remote sensing of environment, 1997, 59(3):440-451. doi: 10.1016/S0034-4257(96)00112-5 [18] LIU H Q, HUETE A R. A feedback based modification of the NDVI to minimize canopy background and atmospheric noise[J]. IEEE Transactions on Geoscience and Remote Sensing, 1995, 33(2):457-465. doi: 10.1109/36.377946 [19] 李文梅, 覃志豪, 李文娟, 等. MODIS NDVI与MODIS EVI的比较分析[J].遥感信息, 2010(6):73-78. http://www.wanfangdata.com.cn/details/detail.do?_type=perio&id=ygxx201006016 [20] 王正兴, 刘闯, 陈文波, 等. MODIS增强型植被指数EVI与NDVI初步比较[J].武汉大学学报(信息科学版), 2006, 31(5):407-410, 427. http://www.cnki.com.cn/Article/CJFDTotal-WHCH200605008.htm [21] CHEN P Y, FEDOSEJEVS G, TISCAREÑO-LÓPEZ M, et al. Assessment of MODIS-EVI, MODIS-NDVI and VEGETATION-NDVI composite data using agricultural measurements:an example at corn fields in western Mexico[J]. Environmental monitoring and assessment, 2006, 119(1-3):69-82. doi: 10.1007/s10661-005-9006-7 [22] ZHOU J, JIA L, MENENTI M.Reconstruction of global MODIS NDVI time series:Performance of Harmonic ANalysis of Time Series(HANTS)[J]. Remote sensing of environment, 2015, 163(Suppl C):217-228. https://www.deepdyve.com/lp/elsevier/reconstruction-of-global-modis-ndvi-time-series-performance-of-bll2kgUmg0 [23] 侯光雷, 张洪岩, 王野乔, 等.基于时间序列谐波分析的东北地区耕地资源提取[J].自然资源学报, 2010, 25(9):1607-1617. doi: 10.11849/zrzyxb.2010.09.020 [24] 梁守真, 邢前国, 施平, 等.山东省典型地表覆被NDVI时间序列谐波分析[J].生态学杂志, 2011, 30(1):59-65. http://www.oalib.com/paper/4972647 -
点击查看大图
图(6) / 表(2)
计量
- 文章访问数: 1193
- HTML全文浏览量: 161
- PDF下载量: 25
- 被引次数: 0
