JECAM JECAMGEOGLAM Science Meeting Brussels Belgium 16 17 November 2015 Investigators Dr Chi Farn Chen CSRSR NCU Taiwan Mr Horng Yuh Gua TARI Taiwan Dr Nguyen Thanh Son CSRSR NCU ID: 819386
Download Pdf The PPT/PDF document "Taiwan TARI site" is the property of its rightful owner. Permission is granted to download and print the materials on this web site for personal, non-commercial use only, and to display it on your personal computer provided you do not modify the materials and that you retain all copyright notices contained in the materials. By downloading content from our website, you accept the terms of this agreement.
JECAM Taiwan TARI siteJECAM/GEOGLAM Sc
JECAM Taiwan TARI siteJECAM/GEOGLAM Science MeetingBrussels, Belgium16-17 November, 2015Investigators:Dr. Chi-Farn Chen, CSRSR, NCU, TaiwanMr. Horng-Yuh
Gua, TARI, TaiwanDr. Nguyen Thanh Son,
Gua, TARI, TaiwanDr. Nguyen Thanh Son, CSRSR, NCU, TaiwanDr. Cheng-Ru Chen, CSRSR, NCU, TaiwanMr. Tsz Feng Lin, TARI, TaiwanSite descriptionâ¢Area: 3,070 km
2â¢64 townshipsâ¢Rice calendar:ï
2â¢64 townshipsâ¢Rice calendar:ï1stcrop: Feb â Julï2ndcrop: Jul â Decâ¢Growing cycle: 110 â 130 daysSite descriptionâ¢Topography: Pla
inâ¢Soil type: Silt loamâ¢Drainag
inâ¢Soil type: Silt loamâ¢Drainage class/ irrigation: Moderate âimperfectâ¢Field size: 0.5 â 1.1 ha â¢Climate: Subtropicalmonsoon/ Typhoon/ Dro
ught â¢Agricultural method: Tradition
ught â¢Agricultural method: Traditional farmingCrop identification and crop area estimationâ¢Aerial photo âRice field â¢Optical satellite imageâFORM
OSAT-2, SPOT, Landsat and so on.â¢D
OSAT-2, SPOT, Landsat and so on.â¢Data fusion âSPOT/MODIS, Landsat/ MODIS â¢SAR âRadarsat-2 2014 (SOAR-JECAM Project)SVM classification result wi
th Lansat 8/MODIS data fusion in 2013S
th Lansat 8/MODIS data fusion in 2013SAR data: Radarsat-2â¢Provided by SOAR-JECAM project â¢1 February~15 October,2014 â¢Mode:VV + HVâ¢1strice
cropping season (Feb.~ Jul.)â¢Res. :
cropping season (Feb.~ Jul.)â¢Res. : 5 mRice backscattering characteristicsNormalized Difference Singma-0 IndexClassification resultRiceNon-riceTotalRef
erenced datarice8611391000non-r
erenced datarice8611391000non-rice1068941000total96710332000Producer Accuracy (%)86.189.3User Accuracy (%)89.086.5Overall Accuracy (%)87.7K
appa coefficient0.75Threshold-base
appa coefficient0.75Threshold-based classification results of the 1stcrop in 2014Image classification resultRice referenced mapIn situ Dataâ¢Land cover
map â¢Soil databaseâ¢Government st
map â¢Soil databaseâ¢Government statistical dataâ¢Yield samplingRecord formrecordCropExcel join to shapefileLand cover map of thecentral-west part
of TaiwanSand micro-photoCharacteris
of TaiwanSand micro-photoCharacteristic CurveWet/dry colour12Soil database derived from soil survey programs Yield surveySun DriedSample DW(Ï)Sampling
Harvesting9 hills of 1 SampleThrashin
Harvesting9 hills of 1 SampleThrashingWeighing5 temporal radar images (Radarsat-2)PreprocessingMultiple regression analysisGeometry correction for investiga
ting dataR² = 0.785450060070080090
ting dataR² = 0.78545006007008009001000110012001300600800100012001400Yield EstimationGround TrustYield Estimation Accuracy 85-90%Prediction of pa
ddy rice yield by multi-temporal rada
ddy rice yield by multi-temporal radar imagesEstimating the yield of a townshipResearch plan and collaborationâ¢Use UAV , VGIS (volunteer geographic informati
onsystem) and IOT/sensor network to co
onsystem) and IOT/sensor network to collect field data.â¢Method validation.â¢New data: ALOS 2, Sentinel 1/ 2 â¢Crop yield model comparisonâ¢The methods
are still improved for mapping other s
are still improved for mapping other short-term crops.â¢We have participated in a collaborative project with scientists from other study sites.Thank youfor