Google Earth Engine, Dữ liệu vệ tinh truy cập mở, và Máy học hỗ trợ lập bản đồ xác suất đầm lầy trên diện rộng

Remote Sensing - Tập 9 Số 12 - Trang 1315
Jennifer Hird1, Evan R. DeLancey2, Gregory J. McDermid1, Jahan Kariyeva2
1Department of Geography, University of Calgary, Calgary, AB T2N 1N4, Canada
2Alberta Biodiversity Monitoring Institute, Edmonton, AB T6G 2E9, Canada

Tóm tắt

Các tiến bộ hiện đại trong điện toán đám mây và các thuật toán máy học đang thay đổi cách sử dụng dữ liệu quan sát Trái Đất (EO) để giám sát môi trường, đặc biệt là trong thời kỳ dữ liệu vệ tinh truy cập mở và miễn phí đang trở nên phổ biến. Việc phân định đầm lầy là một ứng dụng đặc biệt có giá trị của xu hướng nghiên cứu nổi lên này, vì đầm lầy là một thành phần quan trọng về sinh thái nhưng lại thường bị thiếu đại diện trong các chương trình lập bản đồ và giám sát hiện đại, đặc biệt ở cấp độ khu vực và quốc gia. Lợi dụng Google Earth Engine và phần mềm thống kê R, chúng tôi đã phát triển một quy trình công việc để dự đoán khả năng xuất hiện đầm lầy sử dụng mô hình máy học cây hồi quy tăng cường được áp dụng cho dữ liệu địa hình số và EO. Nghiên cứu tại khu vực 13.700 km2 ở Bắc Alberta, mô hình tốt nhất của chúng tôi đã cho ra kết quả xuất sắc, với giá trị AUC (diện tích dưới đường cong đặc tính hoạt động của máy thu) là 0.898 và giá trị sự biến thiên giải thích là 0.708. Kết quả của chúng tôi chứng tỏ vai trò trung tâm của các biến địa hình chất lượng cao trong việc mô hình hóa phân bố đầm lầy ở quy mô khu vực. Việc bao gồm các biến quang học và/hoặc radar vào quy trình đã cải thiện đáng kể hiệu suất của mô hình, mặc dù dữ liệu quang học hoạt động tốt hơn một chút. Việc chuyển đổi mô hình khả năng xuất hiện đầm lầy của chúng tôi thành phân loại nhị phân Wet-Dry cho độ chính xác tổng thể 85%, gần như giống với giá trị thu được từ giải pháp Tổng hợp Đầm lầy Alberta (AMWI): bản kiểm kê đương đại được Chính phủ Alberta sử dụng. Tuy nhiên, quy trình công việc của chúng tôi chứa đựng một số lợi thế chính so với quy trình được sử dụng để sản xuất AMWI, và cung cấp một nền tảng có thể mở rộng cho các sáng kiến giám sát toàn tỉnh.

Từ khóa

#Điện toán đám mây #Máy học #Dữ liệu quan sát Trái Đất #Phân định đầm lầy #Google Earth Engine #Hồi quy tăng cường #Alberta #Vệ tinh truy cập mở #Mô hình hóa đầm lầy #Biến địa hình #Dữ liệu quang học #Dữ liệu radar

Tài liệu tham khảo

Woodcock, 2008, Free access to landsat imagery, Science, 320, 1011, 10.1126/science.320.5879.1011a

Drusch, 2012, Sentinel-2: ESA’s optical high-resolution mission for GMES operational services, Remote Sens. Environ., 120, 25, 10.1016/j.rse.2011.11.026

(2017, May 29). Google. A Planetary-Scale Platform for Earth Science Data & Analysis. Available online: https://earthengine.google.com/.

(2016, September 12). National Aeronautics and Space Administration Welcome to the NASA Earth Exchange (NEX), Available online: https://nex.nasa.gov/nex/.

Amazon Web Services Inc. (2017, November 28). Earth on AWS: Build Planetary-Scale Applications in the Cloud with Open Geospatial Data. Available online: https://aws.amazon.com/earth/.

Chandrashekar, S. (2017, September 12). Announcing Real-Time Geospatial Analytics in Azure Stream Analytics. Available online: https://azure.microsoft.com/en-us/blog/announcing-real-time-geospatial-analytics-in-azure-stream-analytics/.

Yang, 2017, Utilizing Cloud Computing to address big geospatial data challenges, Comput. Environ. Urban Syst., 61, 120, 10.1016/j.compenvurbsys.2016.10.010

Warren, M.S., Brumby, S.P., Skillman, S.W., Kelton, T., Wohlberg, B., Mathis, M., Chartrand, R., Keisler, R., and Johnson, M. (2015, January 13–15). Seeing the Earth in the Cloud: Processing one petabyte of satellite imagery in one day. Proceedings of the 2015 IEEE Applied Imagery Pattern Recognition Workshop (AIPR), Washington, DC, USA.

Hansen, 2013, High-resolution global maps of 21st-century forest cover change, Science, 342, 850, 10.1126/science.1244693

Pekel, 2016, High-resolution mapping of global surface water and its long-term changes, Nature, 540, 418, 10.1038/nature20584

Yamazaki, 2016, The dynamics of Earth’s surface water, Nature, 540, 348, 10.1038/nature21100

DeLancey, E.R., Kariyeva, J., Cranston, J., and Brisco, B. (2017). Monitoring hydro temporal variability in Alberta, Canada with multi-temporal Sentinel-1 SAR data. Can. J. Remote Sens., in press.

Moody, D.I., Warren, M.S., Skillman, S.W., Chartrand, R., Brumby, S.P., Keisler, R., Kelton, T., and Mathis, M. (2016, January 6–9). Building a living Atlas of the earth in the cloud. Proceedings of the 50th Asilomar Conference on Signals, Systems and Computers, Pacific Grove, CA, USA.

Goldblatt, R., You, W., Hanson, G., and Khandelwal, A.K. (2016). Detecting the boundaries of urban areas in India: A dataset for pixel-based image classification in google earth engine. Remote Sens., 8.

Zhou, 2016, ROSCC: An efficient remote sensing observation-sharing method based on cloud computing for soil moisture mapping in precision agriculture, IEEE J. Sel. Top. Appl. Earth Obs. Remote Sens., 9, 5588, 10.1109/JSTARS.2016.2574810

Huntington, J.L., Hegewisch, K.C., Daudert, B., Morton, C.G., Abatzoglou, J.T., McEvoy, D.J., and Erickson, T. (2017). Climate Engine: Cloud computing and visualization of climate and remote sensing data for advanced natural resource monitoring and process understanding. Bull. Am. Meteorol. Soc.

Camps-Valls, G., and Bruzzone, L. (2009). Machine learning techniques in remote sensing data analysis. Kernel Methods for Remote Sensing Data Analysis, John Wiley & Sons.

Richards, 2005, Analysis of remotely sensed data: The formative decades and the future, IEEE Trans. Geosci. Remote Sens., 43, 422, 10.1109/TGRS.2004.837326

Lary, 2016, Machine learning in geosciences and remote sensing, Geosci. Front., 7, 3, 10.1016/j.gsf.2015.07.003

Bey, A., Sánchez-Paus Díaz, A., Maniatis, D., Marchi, G., Mollicone, D., Ricci, S., Bastin, J.-F., Moore, R., Federici, S., and Rezende, M. (2016). Collect Earth: Land use and land cover assessment through augmented visual interpretation. Remote Sens., 8.

Azzari, G., and Lobell, D.B. (2017). Landsat-based classification in the cloud: An opportunity for a paradigm shift in land cover monitoring. Remote Sens. Environ.

Ozesmi, 2002, Satellite remote sensing of wetlands, Wetl. Ecol. Manag., 10, 381, 10.1023/A:1020908432489

Corcoran, 2011, The integration of optical, topographic, and radar data for wetland mapping in northern Minnesota, Can. J. Remote Sens., 37, 564, 10.5589/m11-067

Gabrielsen, 2016, Using a multiscale, probabilistic approach to identify spatial-temporal wetland gradients, Remote Sens. Environ., 184, 522, 10.1016/j.rse.2016.07.034

Maxa, 2009, Mapping northern wetlands with high resolution satellite images and LiDAR, Wetlands, 29, 248, 10.1672/08-91.1

Alberta Environment and Sustainable Resource Development (2013). Alberta Wetland Policy, Alberta Environment and Sustainable Resource Development.

Alberta Environment and Parks (2017, May 29). Alberta Merged Wetland Inventory. Available online: https://geodiscover.alberta.ca/geoportal/catalog/main/home.page.

Kloiber, 2015, A semi-automated, multi-source data fusion update of a wetland inventory for East-Central Minnesota, USA, Wetlands, 35, 335, 10.1007/s13157-014-0621-3

Moore, 1991, Digital terrain modeling : A review of hydrological geomorphological and biological applications, Hydrol. Process., 5, 3, 10.1002/hyp.3360050103

Albaladejo, 2000, Spatial patterns and temporal stability of soil moisture across a range of scales in a semi-arid environment, Hydrol. Process., 14, 1261, 10.1002/(SICI)1099-1085(200005)14:7<1261::AID-HYP40>3.0.CO;2-D

Hogg, 2007, Automated discrimination of upland and wetland using terrain derivatives, Can. J. Remote Sens., 33, S68, 10.5589/m07-049

Devito, 2005, A framework for broad-scale classification of hydrologic response units on the Boreal Plain: Is topography the last thing to consider?, Hydrol. Process., 19, 1705, 10.1002/hyp.5881

Sass, 2008, Characterizing hydrodynamics on boreal landscapes using archived synthetic aperture radar imagery, Hydrol. Process., 22, 1687, 10.1002/hyp.6736

Natural Regions Committee (2006). Natural Regions and Subregions of Alberta, Natural Regions Committee.

Endres, 2017, Mapping boreal peatland ecosystem types from a fusion of multi-temporal radar and optical satellite imagery, Can. J. For. Res., 559, 545

Alberta Environment and Sustainable Resource Development (2015). Alberta Wetland Classification System, Water Policy Branch, Policy and Planning Division.

Smith, 2003, The Forest Watershed and Riparian Disturbance study: A multi-discipline initiative to evaluate and manage watershed disturbance on the Boreal Plain of Canada, J. Environ. Eng. Sci., 2, S1, 10.1139/s03-030

Gorelick, N., Hancher, M., Dixon, M., Ilyushchenko, S., Thau, D., and Moore, R. (2016). Google earth engine: Planetary-scale geospatial analysis for everyone. Remote Sens. Environ.

Weiss, A. (2001, January 9–13). Topographic position and landforms analysis. Proceedings of the Poster Presentation, ESRI User Conference, San Diego, CA, USA.

Alexander, 2016, Micro-topography driven vegetation patterns in open mosaic landscapes, Ecol. Indic., 60, 906, 10.1016/j.ecolind.2015.08.030

Bourgeois, 2013, Application of the topographic position index to heterogeneous landscapes, Geomorphology, 186, 39, 10.1016/j.geomorph.2012.12.015

Wilson, J.P., and Gallant, J.C. (2000). Primary topographic attributes. Terrain Analysis: Principles and Applications, Wiley.

Laamrani, 2015, Distinguishing and mapping permanent and reversible paludified landscapes in Canadian black spruce forests, Geoderma, 237, 88, 10.1016/j.geoderma.2014.08.011

Lang, 2013, Topographic metrics for improved mapping of forested wetlands, Wetlands, 33, 141, 10.1007/s13157-012-0359-8

Beven, 1979, A physically based, variable contributing area model of basin hydrology, Hydrol. Sci. Bull., 24, 43, 10.1080/02626667909491834

Google (2017, May 29). Sentinel-2: MultiSpectral Instrument (MSI), Level-1C. Available online: https://explorer.earthengine.google.com/#detail/COPERNICUS%2FS2.

Rouse, J.W., Haas, R.H., Schell, J.A., and Deering, D.W. (1974). Monitoring Vegetation Systems in the Great Plains with ERTS, Paper-A20; National Aeronautics and Space Administration (NASA).

McFeeters, 1996, The use of the Normalized Difference Water Index (NDWI) in the delineation of open water features, Int. J. Remote Sens., 17, 1425, 10.1080/01431169608948714

Fensholt, 2009, Evaluation of earth observation based long term vegetation trends—Intercomparing NDVI time series trend analysis consistency of Sahel from AVHRR GIMMS, Terra MODIS and SPOT VGT data, Remote Sens. Environ., 113, 1886, 10.1016/j.rse.2009.04.004

Adam, 2010, Multispectral and hyperspectral remote sensing for identification and mapping of wetland vegetation: A review, Wetl. Ecol. Manag., 18, 281, 10.1007/s11273-009-9169-z

Wu, 2014, An effective method for detecting potential woodland vernal pools using high-resolution LiDAR data and aerial imagery, Remote Sens., 6, 11444, 10.3390/rs61111444

Tang, 2016, Assessing Nebraska playa wetland inundation status during 1985–2015 using Landsat data and Google Earth Engine, Environ. Monit. Assess., 188, 654, 10.1007/s10661-016-5664-x

Du, 2016, Water bodies’ mapping from Sentinel-2 imagery with Modified Normalized Difference Water Index at 10-m spatial resolution produced by sharpening the swir band, Remote Sens., 8, 354, 10.3390/rs8040354

European Space Agency (2017, May 29). The SENTINEL-1 Toolbox. Available online: https://sentinel.esa.int/web/sentinel/toolboxes/sentinel-1.

Ulaby, F.T., Moore, R.K., and Fung, A.K. (1986). Microwave Remote Sensing Active and Passive-Volume III: From Theory to Applications, Artech House, Inc.

Patel, 2006, Comparative evaluation of the sensitivity of multi-polarized multi-frequency SAR backscatter to plant density, Int. J. Remote Sens., 27, 293, 10.1080/01431160500214050

Kornelsen, 2013, Advances in soil moisture retrieval from synthetic aperture radar and hydrological applications, J. Hydrol., 476, 460, 10.1016/j.jhydrol.2012.10.044

Mattia, 1997, The effect of surface roughness on multifrequency polarimetric SAR data, IEEE Trans. Geosci. Remote Sens., 35, 954, 10.1109/36.602537

Gherboudj, 2011, Soil moisture retrieval over agricultural fields from multi-polarized and multi-angular RADARSAT-2 SAR data, Remote Sens. Environ., 115, 33, 10.1016/j.rse.2010.07.011

Becker, 1988, Relative sensitivity of normalized difference vegetation Index (NDVI) and microwave polarization difference Index (MPDI) for vegetation and desertification monitoring, Remote Sens. Environ., 24, 297, 10.1016/0034-4257(88)90031-4

Chauhan, 2016, Comparative evaluation of the sensitivity of multi-polarised sar and optical data for various land cover, Int. J. Adv. Remote Sens. Gis Geogr., 4, 1

European Space Agency (2017, November 21). SENTINEL-1 Observation Scenario. Available online: https://sentinel.esa.int/web/sentinel/missions/sentinel-1/observation-scenario.

Pamaploni, P., Marcelloni, G., Paloscia, S., and Sigismondi, S. (1997, January 14–21). The potential of C- and L- band SAR in assessing vegetation biomass: The Ers-1 and JERS-1 experiments. Proceedings of the 3rd ERS Symposium on Space at the Service of Our Environment, Florence, Italy.

Baghdadi, 2001, Evaluation of C-band SAR data for wetlands mapping, Int. J. Remote Sens., 22, 71, 10.1080/014311601750038857

Pope, 1997, Detecting seasonal cycle of the Yucatan Peninsula with SIR-C polarmetric radar imagery, Remote Sens. Environ., 59, 157, 10.1016/S0034-4257(96)00151-4

(2005). Alberta Vegetation Inventory Interpretation Standards, Resource Information Management Branch, Alberta Sustainable Resource Development.

Ducks Unlimited Canada (2011). Enhanced Wetland Classification Inferred Products User Guide, Ducks Unlimited Canada. Version 1.0.

2007, Boosted regression trees for ecological modeling and prediction, Ecology, 88, 243, 10.1890/0012-9658(2007)88[243:BTFEMA]2.0.CO;2

Elith, 2008, A working guide to boosted regression trees, J. Anim. Ecol., 77, 802, 10.1111/j.1365-2656.2008.01390.x

Buston, 2011, Determinants of reproductive success in dominant pairs of clownfish: A boosted regression tree analysis, J. Anim. Ecol., 80, 528, 10.1111/j.1365-2656.2011.01803.x

Parisien, 2011, Scale-dependent controls on the area burned in the boreal forest of Canada, 1980–2005, Ecol. Appl., 21, 789, 10.1890/10-0326.1

Parisien, 2014, An analysis of controls on fire activity in boreal Canada: Comparing models built with different temporal resolutions, Ecol. Appl., 24, 1341, 10.1890/13-1477.1

(2016). R Development Core Team R: A Language and Environment for Statistical Computing, R Foundation for Statistical Computing.

Ridgeway, G. (2017, December 08). GBM: Generalized Boosted Regression Models. Available online: https://cran.r-project.org/web/packages/gbm/gbm.pdf.

Swets, 1988, Measuring the accuracy of diagnostic systems, Science, 240, 1285, 10.1126/science.3287615

Zweig, 1993, Receiver-operating characteristic (ROC) plots: A fundamental evaluation tool in clinical medicine, Clin. Chem., 39, 561, 10.1093/clinchem/39.4.561

Freeman, 2008, A comparison of the performance of threshold criteria for binary classification in terms of predicted prevalence and kappa, Ecol. Model., 217, 48, 10.1016/j.ecolmodel.2008.05.015

Guisan, 2000, Predictive habitat distribution models in ecology, Ecol. Model., 135, 147, 10.1016/S0304-3800(00)00354-9

Allouche, 2006, Assessing the accuracy of species distribution models: Prevalence, kappa and the true skill statistic (TSS), J. Appl. Ecol., 43, 1223, 10.1111/j.1365-2664.2006.01214.x

Congalton, 1991, A review of assessing the accuracy of classifications of remotely sensed data, Remote Sens. Environ., 37, 35, 10.1016/0034-4257(91)90048-B

Murphy, 2007, Mapping wetlands: A comparison of two different approaches for New Brunswick, Canada, Wetlands, 27, 846, 10.1672/0277-5212(2007)27[846:MWACOT]2.0.CO;2

Lidberg, 2014, Evaluating digital terrain indices for soil wetness mapping-a Swedish case study, Hydrol. Earth Syst. Sci., 18, 3623, 10.5194/hess-18-3623-2014

Hogg, 2008, An evaluation of DEMs derived from LiDAR and photogrammetry for wetland mapping, For. Chron., 84, 840, 10.5558/tfc84840-6

Riley, 2017, Identifying small depressional wetlands and using a topographic position index to infer hydroperiod regimes for pond-breeding amphibians, Wetlands, 37, 325, 10.1007/s13157-016-0872-2

Kasischke, 1997, The use of imaging radars for ecological applications—A review, Remote Sens. Environ., 59, 141, 10.1016/S0034-4257(96)00148-4

Government of Canada (2017, November 27). Historical Climate Data. Available online: http://climate.weather.gc.ca/index_e.html.

Alberta Agriculture and Forestry (2017, November 27). Current and Historical Alberta Weather Station Data Viewer. Available online: https://agriculture.alberta.ca/acis/alberta-weather-data-viewer.jsp.

(2017, October 02). Alberta Biodiversity Monitoring Institute. 3 × 7-km Photoplot Land Cover Data. Available online: http://abmi.ca/home/data-analytics/da-top/da-product-overview/GIS-Human-Footprint-Land-Cover-Data/Photoplot-Land-Cover-Dataset.html.

(2017, October 02). Alberta Biodiversity Monitoring Institute. 3 × 7-km Sample-Based Human Footprint Data. Available online: http://abmi.ca/home/data-analytics/da-top/da-product-overview/GIS-Human-Footprint-Land-Cover-Data/Human-Footprint-Sample-Based-Inventory.html.

European Space Agency (2017, November 27). The Sentinel-2 Toolbox. Available online: https://sentinel.esa.int/web/sentinel/toolboxes/sentinel-2.

(2017, December 08). Alberta Environment and Parks. Available online: http://aep.alberta.ca/forms-maps-services/maps/resource-data-product-catalogue/biophysical.aspx.