Monitoring Vegetation Water Content by Using Optical Vegetation
Monitoring Vegetation Water Content by Using Optical Vegetation Index and Microwave Vegetation Index: Field Experiment and Application Hui Lu ( Tsinghua University, China) Toshio Koike & Hiroyuki Tsutsui (The University of Tokyo) Hedeyuki Fujii (JAXA) IGARSS 2011, Jul. 27, Vancouver 1
Outline • Background and motivation • Microwave vegetation index • Field Experiment – Setting and instruments – Observed Results • Application – Mongolia site • Remark IGARSS 2011, Jul. 27, Vancouver 2
Background • Global vegetation information is closely related to – Food productivity, famine, …… – Environment, ecological system, …. • In land surface modeling and remote sensing retrieval, vegetation is – A key variable of land surface remote sensing • Soil moisture, soil temperature, vegetation water content – A key parameter in GCM, hydrology and land surface scheme • LAI, f. PAR, ET, precipitation interception – A key parameter in terrestrial ecosystem model • Carbon cycle IGARSS 2011, Jul. 27, Vancouver 3
Motivation • Vegetation parameters observed by satellites: – VIS/IR: fractional coverage, NDVI, LAI, NDWI, EVI – MW: Vegetation water content (VWC), Microwave vegetation index (MVI) • MW RS has daily global coverage and deeper penetration depth – Complement vegetation information to VIS/IR • What the relationship between these parameters? • Accurate VWC is useful in – Improving soil moisture retrieval algorithm – Improving LDAS IGARSS 2011, Jul. 27, Vancouver 4
VWC, MVI, NDWI • Microwave vegetation index by Shi • NDVI: VIS (620 - 670 nm) & NIR (841 - 876 nm) • NDWI: SWIR in band 5 (1230 -1250 nm) or band 6 (1628 -1652 nm) IGARSS 2011, Jul. 27, Vancouver 5
Field Experiment --Instruments and setting l. Brightness temperature observed by Ground Based Microwave Radiometer, at 6. 925, 10. 65, 18. 7, 23. 8, 36. 5, 89 GHz l. VIS/IR reflectance measured by ASD Field. Spec Pro in a spectral range of 350 nm – 2500 nm IGARSS 2011, Jul. 27, Vancouver 6
Experiment design 3 2 IGARSS 2011, Jul. 27, Vancouver 1 7
Experiment design • Observing winter wheat – One kind of main crops – VWC is not so big, C-band could penetrate. Winter wheat development VWC was measured by sampling IGARSS 2011, Jul. 27, Vancouver 8
Vegetation 11 -29 01 -16 12 -08 01 -19 12 -13 01 -24 IGARSS 2011, Jul. 27, Vancouver 12 -20 02 -07 9
Observed Results VWC ~ NDVI shows a poor correlation to the VWC, with an R-square less than 0. 2. It is not good to estimate VWC from NDVI observation! IGARSS 2011, Jul. 27, Vancouver 10
Observed Results VWC ~ NDWI has a good correlation to VWC, while band 5 has bigger R value VWC information maybe can be estimated by NDWI 5, for vwc in [0, 4] IGARSS 2011, Jul. 27, Vancouver 11
Observed Results VWC ~ MWI High R for X-C band VWC = linear regression function of MVI IGARSS 2011, Jul. 27, Vancouver 12
Application Domain • AMPEX – Mongolia; – Relative homogenous – VWC survey at 2003 Jul and Aug; – 160*120 km; IGARSS 2011, Jul. 27, Vancouver 13
Application: VWC retrieved from JAXA algorithm Vs. in situ l VWC provided by JAXA algorithm is comparable to the in situ observed VWC l. Using as reference data to check the performance of MVI-based method IGARSS 2011, Jul. 27, Vancouver 14
Results: VWC from MVI-based method MVI(10, 6) MVI (18, 10) A 3 H 7 High R for X-C band IGARSS 2011, Jul. 27, Vancouver 15
Remark • Field experiment which observing winter wheat development by using microwave radiometer and VIS/IF spectroradiometer simultaneously. • Comparing to in situ observed VWC – NDVI show poor correlation – NDWI show good correlation – MVI show strong correlation • MVI-based linear equation could provide VWC information, but the absolute values should be scaled – Can be used to monitor the vegetation temporal variation – The coefficient of linear equation should be related to (vfc, vegetation type) • Future work: – Quantify the coefficient by each vegetation type (LSM classification, or real type) – Test for more observation sites (US site, MDB site, China) IGARSS 2011, Jul. 27, Vancouver 16
Thank you for your attention! IGARSS 2011, Jul. 27, Vancouver 17
- Slides: 17