亚洲区域AMSR2与SMOS土壤水分产品对比研究

doi:10.11873/j.issn.1004-0323.2019.1.0125

鎽樿/Abstract

鎽樿锛� 閬ユ劅鍙嶆紨鍦熷￥姘村垎(SM)浜у搧瓒婃潵瓒婂鍦板簲鐢ㄤ簬鍐滀笟銆佹皵璞°�佹按鏂囩瓑鐮旂┒锛岃�屽井娉㈠湡澹ゆ按鍒嗘暟鎹骇鍝佺殑鍖哄煙閫傜敤鎬у垎鏋愭槸鍏跺悎鐞嗕娇鐢ㄧ殑蹇呰鍓嶆彁銆備娇鐢∕ERRA-2(Modern Era Retrospective-analysis for Research and Applications锛孷ersion 2)妯℃嫙鍦熷￥姘村垎涓哄弬鑰冩暟鎹紝杩愮敤浼犵粺缁熻鏂规硶(鍘熷鏁版嵁鐩稿叧鎬с�佽窛骞崇浉鍏虫�с�佸亸宸互鍙婃棤鍋忓潎鏂规牴宸�)鍜孴C(Triple-Collocation)涓嶇‘瀹氭�ц宸ā鍨嬪垎鏋愮殑鏂规硶锛屽浜氭床鍖哄煙2012骞�7鏈垀2016骞�7鏈堜袱绉嶈鍔ㄥ井娉㈠湡澹ゆ按鍒哠MOS-L3-SM(Soil Moisture and Ocean Salinity锛孡3)鍜孉MSR2-LPRM-SM(The Advanced Microwave Scanning Radiometer 2锛孡and Parameter Retrieval Model Product)杩涜瀵规瘮璇勪及銆傜粨鏋滆〃鏄庯細鈶犵┖闂翠笂SMOS-L3杈傾MSR2-LPRM鏁版嵁涓庡弬鑰冩暟鎹甅ERRA-2鍦熷￥姘村垎鐨勭浉鍏虫�ц緝濂斤紝琛ㄧ幇涓篠MOS-L3-SM鍏锋湁杈冨ソ鐨勭┖闂磋繛缁�э紝涓斿湪浜氭床澶у鏁板湴鍖烘湁杈冨皬鐨勬棤鍋忓潎鏂规牴宸�;鈶℃箍瀛ｆ潯浠朵笅閬ユ劅鍦熷￥姘村垎涓庡弬鑰冨�肩殑鐩稿叧鎬ф瘮骞插鏉′欢涓嬬殑鐩稿叧鎬ф洿濂斤紝涓斿共瀛ｅ嚭鐜伴珮绾湴鍖�(绾�>55掳)缂哄け鍊艰緝澶氱殑鎯呭喌;鈶袱閬ユ劅鍦熷￥姘村垎鐨凾C璇樊鍛堢幇鐩镐技鐨勫垎甯冿紝鍖哄煙TC骞冲潎璇樊涓よ�呭潎涓�0.076 m³/m³顎夈�傛�讳箣锛孲MOS-L3-SM鍜孉MSR2-LPRM-SM鍦ㄧ┖闂寸浉鍏虫�у強TC璇樊璇勪环鏂归潰閮藉叿鏈夊悎鐞嗘�э紝涓洪仴鎰熷湡澹ゆ按鍒嗗湪鍐滀笟銆佹皵璞°�佹按鏂囩瓑鏂归潰鐨勫簲鐢ㄦ彁渚涘弬鑰冦��

鍏抽敭璇�: 鍦熷￥姘村垎, SMOS-L3-SM, AMSR2-LPRM-SM, MERRA-2, 浜氭床

Abstract:

Soil Moisture (SM)products derived from satellite missions have been widely applied in agriculture锛宮eteorology锛宧ydrology锛宎nd other fields.It is very necessary to assess the reliability of satellite soil moisture products over regional scale before using them.Two passive satellite soil moisture products锛宱btained from SMOS (the Soil Moisture Ocean Salinity锛汱3)and AMSR2 (the Advanced Microwave Scanning Radiometer 2锛汱PRM)锛寃ere evaluated over Asia with reference to MERRA-2 (the Modern Era Retrospective-analysis data for Research and Applications锛孷ersion 2)simulated products.The evaluation was performed by adopting a classic statistical method (including Pearson correlation (R)for original SM data and anomalies锛宐ias and unbiased root mean square root)and Triple Collocation (TC)approach based on the SMOS-L3 and AMSR-LPRM daily SM products during July 2012~July 2016.The results reveal that锛�(1)in space锛宼he performance of SMOS-L3-SM is better than AMSR2-LPRM-SM both for the original SM data and anomalies.Because the SMOS-L3-SM performed consistent correlation over Asia and the smaller unbiased root mean squared difference (ubRMSD)of SMOS-L3-SM is found in most parts of Asia锛�(2)the correlation between satellite and simulated SM is better in the wet season than in the dry season.Additionally there is a quite high probability of a lack of value in high latitude (about >55掳)areas in the dry season to appear锛�(3)SMOS-L3-SM and AMSR2-LPRM-SM have relatively similar TC errors distribution with a mean error of 0.076 m³/m³for both of them.Overall锛宐oth SMOS-L3-SM and AMSR2-LPRM-SM give reasonable results on correlation and TC error锛宼hus providing an additional reference for application of soil moisture in agriculture锛宮eteorology锛宧ydrology and other studies.

Key words: Soil moisture锛汼MOS-L3-SM锛汚MSR2-LPRM-SM, MERRA-2, Asia

涓浘鍒嗙被鍙�:

TP75

鏉庣憺濞�, 鏉庡厗瀵�, 閮濈澘, 寮犺垝鏄�, 娼樺墤鍚�. 浜氭床鍖哄煙AMSR2涓嶴MOS鍦熷￥姘村垎浜у搧瀵规瘮鐮旂┒[J]. 閬ユ劅鎶�鏈笌搴旂敤, 2019, 34(1): 125-134.

Li Ruijuan, Li Zhaofu, Hao Rui, Zhang Shuyu, Pan Jianjun. A Regional-Scale Performance Evaluation of SMOS and AMSR2 Soil Moisture Products over Asia[J]. Remote Sensing Technology and Application, 2019, 34(1): 125-134.

鍙傝�冩枃鐚�

锛�1锛絎ang Q W锛孏u Y F锛孴uia D.Discriminative Multiple Kernel Learning for Hyperspectral Image Classification[J].IEEE Transactions on Geoscience and Remote Sensing锛�2016锛�54(7)锛�1-16 [2]Pedro D.A Few Useful Things to Know About Machine Learning[J].Communications of the ACM锛�2012锛�55(10)锛�78-87. [3]Kotsiantis S B锛孼aharakis I D锛孭intelas P E.Machine Learning锛欰 Review of Classification and Combining Techniques[J].Artificial Intelligence Review锛�2006锛�26(3)锛�159-190. [4]Corinna C锛孷ladimir V.Support-Vector Networks[C]鈭� Proc.of the 25th Machine Leaning锛�1995锛�20(3)锛�273-297. [5]Madan S锛孭ranjali C.A Review of Machine Learning Techniques Using Decision Tree and Support Vector Machine[A].International Conference on Computing Communication Control & Automation[C]鈭iscataway锛孨 J锛孶SA锛欼EEE锛�2017锛�1-7. [6]Zhu Q Y锛孮in A K锛孲uganthan P N锛宔t al.Evolutionary Extreme Learning Machine[J].Pattern Recognition锛�2005锛�38(10)锛�1759-1763. [7]Huang G B锛孼hu Q Y锛孲iew C K.Extreme Learning Machine锛歍heory and Applications[J].Neurocomputing.2006锛�70(1)锛�489-501. [8]Dhandhalya J K锛孭armar S K.Hyperspectral Image Classification Using Spatial Spectral Features and Machine Learning Approach[C]鈭EEE International Conference on Recent Trends In Electronics Piscataway锛孨 J锛孶SA锛�2017锛�1161-1165. [9]Chen C锛孡i W锛孲u H锛宔t al.Spectral-spatial Classification of Hyperspectral Image based on Kernel Extreme Learning Machine[J].Remote Sensing锛�2014锛�6(6)锛�5795-5814. [10]Roerdink J锛孧eijster A.The Watershed Transform锛欴efinitions锛孉lgorithms and Parallelization Strategies[J].Fundamenta Informaticae锛�2000锛�41(1锛�2)锛�187-228. [11]Michailovich O锛孯athi Y锛孌olui S.Spatially Regularized Compressed Sensing for High Angular Resolution Diffusion Imaging[J].IEEE Transactions on Medical Imaging锛�2011锛�30(5)锛�1100-1115. [12]Manjunath B S锛孧a W Y.Texture Features for Browsing and Retrieval of Image Data[J].IEEE Trans Pami锛�1996锛�18(8)锛�837-842. [13]Shao Liangshan锛孧a Han锛學en Tingxin.Study on Classifier Ensemble Learning Model for Extreme Learning Machine[J].Computer Engineering and Applications锛�2016锛�52(13)锛�121-125.[閭佃壇鏉夛紝椹瘨锛屾俯寤锋柊.鏋侀檺瀛︿範鏈虹殑鍒嗙被鍣ㄩ泦鎴愭ā鍨嬬爺绌禰J].璁＄畻鏈哄伐绋嬩笌搴旂敤锛�2016锛�52(13)锛�121-125.] [14]Yi Xufeng锛學u Jian锛孻u Gaohang.Parameter Selection of RBF Kernel Using Kernel Polarization by An Adaptive Gradient Method[J].Journal of Gannan Normal University锛�2015锛�36(6)锛� 23-27.[鏄撳簭宄帮紝鍚村潥锛屽柣楂樿埅.鍩轰簬鑷�傚簲姊害绠楁硶鐨凷VM鏍告瀬鍖栧弬鏁伴�夋嫨鏂规硶[J].璧ｅ崡甯堣寖瀛﹂櫌瀛︽姤锛�2015锛�36(6)锛�23-27.] [15]Zhao Zhiyong锛孡i Yuanxiang锛孻u Fei锛宔t al.Improved Deep Learning Algorithm based on Extreme Learning Machine[J].Computer Engineering and Design锛�2015锛�36(4)锛�1022-1026.[璧靛織鍕囷紝鏉庡厓棣欙紝鍠婚锛岀瓑.鍩轰簬鏋侀檺瀛︿範鐨勬繁搴﹀涔犵畻娉昜J].璁＄畻鏈哄伐绋嬩笌璁捐锛�2015锛�36(4锛夛細1022-1026.] [16]Zhang Jing锛孻ao Shuke.The Application of the Data Mining To the Malware Detection[J].Computer Knowledge and Technology锛�2011锛�7(35)锛�9048-9049.[寮犳暚锛屽涔︾.鏁版嵁鎸栨帢鍦ㄦ伓鎰忕▼搴忔娴嬩腑鐨勫簲鐢╗J].鐢佃剳鐭ヨ瘑涓庢妧鏈紝2011锛�7(35)锛�9048-9049] [17]Wei Ao.Design of A Classifier based on the SVM Algorithm[J].Electronic Science and Technology锛�2015锛�28(4)锛�23-26.[鍗卞偛.鍩轰簬SVM绠楁硶鐨勫垎绫诲櫒璁捐[J].鐢靛瓙绉戞妧锛�2015锛�28(4)锛�23-26.] [18]Wang S锛孡iu Q锛孼hu E锛宔t al锛孒yperparameter Selection of One-class Support Vector Machine by Self-adaptive Data Shifting[J].Pattern Recognition锛�2018锛�74锛�198-211. [19]Qiu Yunfei锛孡i Zhiyi.An Improved Artificial Fish Swarm Algorithm and Its Application for SVM Parameter Optimization[J].Computer Engineering and Science锛�2018锛�40(11)锛�2074-2079.[閭变簯椋烇紝鏉庢櫤涔�.鏀硅繘浜哄伐楸肩兢绠楁硶鍦⊿VM鍙傛暟浼樺寲涓殑搴旂敤[J].璁＄畻鏈哄伐绋嬩笌绉戝锛�2018锛�40(11)锛�2074-2079.] [20]Heras D B锛孉rg眉ello F锛孮uesadabarriuso P.Exploring ELM-based Spatial-spectral Classification of Hyperspectral Images[J].International Journal of Remote Sensing锛�2014锛�35(2)锛�401-423. [21]Gao Wenjie锛學ang Jinliang锛孡iu Guangjie.Study on the Effect of Training Sample on Classification Accuracy of Remote Sensing Image[J].Yunnan Geographic Environment Research锛�2015锛�27(2)锛�31-36.[楂樻枃鏉帮紝鐜嬮噾浜紝鍒樺箍鏉�.璁粌鏍锋湰瀵归仴鎰熷奖鍍忓垎绫荤簿搴﹀奖鍝嶇爺绌禰J].浜戝崡鍦扮悊鐜鐮旂┒锛�2015锛�27(2)锛�31-36.] [22]Fandino J L锛孊arriuso P Q锛孒eras D B锛宔t al.Efficient ELM-based Techniques for the Classification of Hyperspectral Remote Sensing Images on Commodity Gpus[J].IEEE Journal of Selected Topics In Applied Earth Observations and Remote Sensing锛�2017锛�8(6)锛�2884-2893. [23]Barriuso P Q锛孉rguello F锛孒eras D B.Spectral-Spatial Classification of Hyperspectral Images Using Wavelets and Extended Morphological Profiles[J].IEEE Journal of Selected Topics In Applied Earth Observations and Remote Sensing锛�2014锛�7(4)锛�1177-1185. [24]Lu Siyuan锛孡u Zhihai锛學ang Shuihua锛宔t al.Review of Extreme Learning Machine[J].Measurement & Control Technology锛�2018锛�37(10)锛�3-9.[闄嗘�濇簮锛岄檰蹇楁捣锛岀帇姘磋姳锛岀瓑.鏋侀檺瀛︿範鏈虹患杩癧J].娴嬫帶鎶�鏈紝2018锛�37(10)锛�3-9.] [25]Huang G B锛孌ing X锛孼hou H.Optimization Method based Extreme Learning Machine for Classification[J].Neurocomputing锛�2010锛�74(1-3)锛�155-163. [26]Huang G B锛孊abri H A.Upper Bounds on the Number of Hidden Neurons in Feedforward Networks With Arbitrary Bounded Nonlinear Activation Functions[J].IEEE Trans Neural Netw锛�1998锛�9(1)锛�224-229. [27]Yu Qiuyan锛孼hao Ying锛孲un Jijia锛宔t al.Realization and Comparison of Typical Machine Learning Algorithms In Fatty Liver Classification Prediction Research[J].Journal of Mathematical Medicine锛�2019锛�32(1)锛�1-3.[浣欑鐕曪紝璧佃幑锛屽瓩缁т匠锛岀瓑.鍏稿瀷鏈哄櫒瀛︿範绠楁硶鍦ㄨ剛鑲倽鍒嗙被棰勬祴鐮旂┒涓殑瀹炵幇涓庢瘮杈僛J].鏁扮悊鍖昏嵂瀛︽潅蹇楋紝2019锛�32(1)锛�1-3.] [28]Cao Weichao锛孡i Ming锛孡iu Jiang锛宔t al.Study on Evaluation of Land Cover Classification Data based on Moderate Resolution Imaging[J].Geomatics & Spatial Information Technology锛�2018锛�41(11)锛�55-57锛�61.[鏇逛紵瓒咃紝鏉庢槑锛屽垬姹燂紝绛�.鍩轰簬涓垎褰卞儚鐨勪笉鍚屾椂鐩稿湴琛ㄨ鐩栧垎绫绘暟鎹簿搴﹁瘎浠穂J].娴嬬粯涓庣┖闂村湴鐞嗕俊鎭紝2018锛�41(11)锛�55-57锛�61.] [29]Wang Hao锛孋hen Yanhua锛孡i Yan.Kappa Analysis of Prescription Comment From Different Audit Directions[J].Practical Pharmacy and Clinical Remedies锛�2018锛�21(10)锛�1200-1202.[姹旦锛岄檲鐕曞崕锛屾潕鐕�.浠庝笉鍚屽鏍告柟鍚戣繘琛屽鏂圭偣璇勭殑Kappa鍒嗘瀽[J].瀹炵敤鑽墿涓庝复搴婏紝2018锛�21(10)锛�1200-1202.] [30]Cai L锛孲hi W锛孧iao Z锛宔t al.Accuracy Assessment Measures for Object Extraction from Remote Sensing Images[J].Remote Sensing锛�2018锛�10(2)锛�303.doi锛�10.3390/rs10020303.

[1]	琚佸惎杩�,璧垫灄,骞歌禐鍝�,閭瑰痉瀵�,璧靛缓濠�,鍒樹笘鍗�,妯婄. 闈掕棌宸ョ▼璧板粖娌跨嚎澶氬勾鍐诲湡鍖鸿瀺鍖栧楂樺垎杈ㄧ巼鍦拌〃鍦熷￥姘村垎鍙嶆紨[J]. 閬ユ劅鎶�鏈笌搴旂敤, 2023, 38(6): 1350-1363.
[2]	鏉ㄥ, 姹ょ嚂鏉�, 寮犲畞棣�, 寮犳亽鏉�, 寰愬皯鍗�. 鍩轰簬SMOS銆丼MAP鏁版嵁鐨勯潚钘忛珮鍘熷椋庡強妞嶈鐢熼暱瀛ｅ湡澹ゆ按鍒嗛暱娑堢壒寰佺爺绌�[J]. 閬ユ劅鎶�鏈笌搴旂敤, 2022, 37(6): 1373-1384.
[3]	鏇归珮宀�,鏉ㄥ. 鍩轰簬鍏抽敭杈呭姪鍙傛暟鏁版嵁鐨凩-MEB浜害娓╁害妯℃嫙鐮旂┒[J]. 閬ユ劅鎶�鏈笌搴旂敤, 2022, 37(6): 1385-1391.
[4]	榛勫仴搴�,鏉ㄥ,椹秴. SMAP L2澶氬昂搴﹀湡澹ゆ按鍒嗘暟鎹笌ISMN瀹炴祴鏁版嵁鐨勫樊寮傜壒寰佺爺绌�[J]. 閬ユ劅鎶�鏈笌搴旂敤, 2022, 37(6): 1392-1403.
[5]	鏉滅粛鏉�,璧靛ぉ鏉�,鏂藉缓鎴�,椹槬閿�,閭瑰痉瀵�,鐜嬫尟,濮氱浖鐩�,褰織鏅�,閮戞櫙鑰�. Sentinel-1鍜孲entinel-2鍗忓悓鍙嶆紨鍦拌〃鍦熷￥姘村垎[J]. 閬ユ劅鎶�鏈笌搴旂敤, 2022, 37(6): 1404-1413.
[6]	榛勯挵鐜�,鍒樺嚡,鐜嬫爲涓�,鐜嬪ぇ鎴�,鑻戝嘲,鐜嬩繚鏋�,鏅枃,鐜嬩紵. 鍖楁柟鍏稿瀷鍖哄煙閬ユ劅鍜屾ā鍨嬪湡澹ゆ按鍒嗕骇鍝佺殑瀵规瘮鍙婅瘎浼�[J]. 閬ユ劅鎶�鏈笌搴旂敤, 2022, 37(6): 1414-1426.
[7]	鐜嬮洩鐞�,寮犵繑,闄堣兘鎴�,椹畯浜�. 鍩轰簬鏃剁┖鍙樺紓鎬х殑鍦熷￥姘村垎瑙傛祴缃戣璁�[J]. 閬ユ劅鎶�鏈笌搴旂敤, 2022, 37(6): 1437-1446.
[8]	鏂硅タ鐟�,钂嬬幉姊�,宕旀収鐝�. 鍩轰簬Sentinel-1闆疯揪鏁版嵁鐨勯潚钘忛珮鍘熷湴鍖哄湡澹ゆ按鍒嗗弽婕旂爺绌�[J]. 閬ユ劅鎶�鏈笌搴旂敤, 2022, 37(6): 1447-1459.
[9]	鏅忕孩娉�,鏉庢旦,鍗㈢尞鍋�,鐜嬩匠鍗�. 鍩轰簬LST-VI鐗瑰緛绌洪棿鐨勫杸鏂壒鍦板尯鍦熷￥姘村垎鏃剁┖鍙樺寲鐮旂┒[J]. 閬ユ劅鎶�鏈笌搴旂敤, 2022, 37(6): 1460-1471.
[10]	搴锋柊绀�,寮犳枃璞�,鍒樺師钀�,椤捐鍙�,浣欐稕,寮犱附涓�,寰愭ˇ鏄�. 鍩轰簬闅忔満妫灄鐨勪含娲ュ唨鍦板尯PM_2.5閬ユ劅鍙嶆紨鍙婂彉鍖栧垎鏋�[J]. 閬ユ劅鎶�鏈笌搴旂敤, 2022, 37(2): 424-435.
[11]	鐜嬬殦,閮濊幑,琚佹澗,闄堝厜鑸�,闈宠帀鑾�. SMAP鍦熷￥姘村垎浜у搧鍦ㄦ樊娌虫祦鍩熺殑閫傜敤鎬ц瘎浼�[J]. 閬ユ劅鎶�鏈笌搴旂敤, 2021, 36(5): 1009-1021.
[12]	濮滄稕,閮戝叴鏄�,鏉庢檽娲�,鏉庢檽宄�,璧靛嚡. 闂數娌虫祦鍩烲娉㈡鍦板熀寰尝杈愬皠璁FI妫�娴嬩笌鐢电鐜鍒嗘瀽[J]. 閬ユ劅鎶�鏈笌搴旂敤, 2021, 36(5): 1022-1032.
[13]	鑲栫獔,鏇捐秴,娌堢剷閿�. 缁撳悎鍙傞噺缁熻涓庢椂绌鸿瀺鍚堢殑鍦熷￥姘村垎闄嶅昂搴︽柟娉�[J]. 閬ユ劅鎶�鏈笌搴旂敤, 2021, 36(5): 1033-1043.
[14]	鏉ㄦ婧�,鐧芥檽闈�. 鍩轰簬Sentinel-1鍜孧ODIS鏁版嵁鍙嶆紨鍐滅敯鍦拌〃鍦熷￥姘村垎鈥斾互REMEDHUS鍦板尯涓轰緥[J]. 閬ユ劅鎶�鏈笌搴旂敤, 2021, 36(5): 973-982.
[15]	鏉ㄥゥ鑾�,閮戜笢娴�,鏂囧啗,闄嗗鎵�,鏉ㄨ秺,绗︽櫞. 闈掕棌楂樺師L娉㈡寰尝杈愬皠瑙傛祴涓庡湡澹ゆ按鍒嗗弽婕旂爺绌惰繘灞�[J]. 閬ユ劅鎶�鏈笌搴旂敤, 2021, 36(5): 983-996.

浜氭床鍖哄煙AMSR2涓嶴MOS鍦熷￥姘村垎浜у搧瀵规瘮鐮旂┒

A Regional-Scale Performance Evaluation of SMOS and AMSR2 Soil Moisture Products over Asia

PDF (PC)

鍙鍖�

鎽樿/Abstract

寮曠敤鏈枃

浣跨敤鏈枃

鍙傝�冩枃鐚�

鐩稿叧鏂囩珷 15

缂栬緫鎺ㄨ崘

Metrics

鏈枃璇勪环