Sensors 2010,10,6980-7001;doi:10.3390/s100706980 O PEN A C C ESS sensors ISSN 1424-8220 w w w m dpi.com /journal/sensors Article L Band Brightness Tem perature O bservations over a C orn C anopy during the Entire G row th C ycle A licia T.Joseph 1,*,R ogier van der V elde 2,Peggy (2¶1HLOO 1,Bhaskar J.C houdhury 1, R oger H Lang 3,Edw ard J.K im and Tim othy G ish 4 H ydrologicalSciences Branch/614.3,H ydrospheric and Biospheric Sciences Laboratory, N A SA /G oddard Space FlightCenter,G reenbelt,M D 20771,U SA ; E-M ails:Peggy.E.O N eill@ nasa.gov (P.E.O );Bhaskar.J.Choudhury@ nasa.gov (B.J.C.); Edw ard.J.K im @ nasa.gov (E.J.K ) InternationalInstitute forG eo-Inform ation Science and Earth O bservation (ITC),H engelsostraat99, P.O Box 6,7500 A A Enschede,The N etherlands;E-M ail:velde@ itc.nl D epartm entofElectricalEngineering & Com puterSciences,the G eorge W ashington U niversity, W ashington,D C 20052,U SA ;E-M ail:lang@ gw u.edu U SD A -A RS H ydrology and Rem ote Sensing Laboratory,Building 007,BA RC-W EST,Beltsville, M D 20705,U SA ;E-M ail:Tim othy.G ish@ ars.usda.gov * A uthorto w hom correspondence should be addressed;E-M ail:A licia.T.Joseph@ nasa.gov; Tel.:+1-301-614-5804;Fax:+1-301-614-5808 Received:26 M arch 2010;in revised form :11 June 2010 /Accepted:11 June 2010 / Published:20 July 2010 A bstract:D uring a field cam paign covering the 2002 corn grow ing season,a dualpolarized tow erm ounted L-band (1.4 G H z)radiom eter(LRA D )provided brightness tem perature (TB ) m easurem ents at preset intervals, incidence and azim uth angles These radiom eter m easurem ents w ere supported by an extensive characterization of land surface variables including soilm oisture,soiltem perature,vegetation biom ass,and surface roughness.In the period M ay 22 to A ugust30,ten days ofradiom eterand ground m easurem ents are available fora corn canopy w ith a vegetation w atercontent(W )range of0.0 to 4.3 kg m í2.U sing this data set,the effects of corn vegetation on surface em issions are investigated by m eans of a sem i-em pirical radiative transfer m odel A dditionally, the im pact of roughness on the surface em ission is quantified using TB m easurem ents over bare soil conditions Subsequently,the estim ated roughness param eters,ground m easurem ents and horizontally Sensors 2010,10 6981 (H )-polarized TB are em ployed to invert the H -polarized transm issivity (Ȗh) for the m onitored corn grow ing season K eyw ords:field cam paign;L-band radiom etry;vegetation effects;surface roughness 1.Introduction Low frequency passive m icrow ave observations have been intensively studied fortheirpotentialof retrieving soil m oisture e.g., [1-3] Studies have dem onstrated that w hen an appropriate characterization of vegetation, soil surface roughness and dielectric properties are applied, soil m oisture can be retrieved fairly accurately from the brightness tem peratures (TB ¶V PHDVXUHG E\ m icrow ave radiom eters e.g., [4,5] A s a result, the Soil M oisture and O cean Salinity (SM O S [6]) m ission is the firstofthree L-band radiom eters designed forglobalsoilm oisture m onitoring purposes to be launched.In the nearfuture,the A quarius and SoilM oisture A ctive Passive (SM A P [7])m issions w ill follow ; their expected launch dates are in spring 2010 and in 2015, respectively W ith this increased availability of low frequency spaceborne radiom eter observations,new opportunities arise form onitoring soilm oisture globally A m ong the challenges in retrieving soil m oisture from TB m easurem ents is to account for soil surface roughness and vegetation effects M ost retrieval approaches utilize sim ilar radiative transfer equations [8-10],in w hich the effects ofvegetation are param eterized by the vegetation transm issivity (Ȗ) and the single scattering albedo (Ȧ) Pardé et al [11] concluded that for retrieving soil m oisture globally the Ȧ can be assum ed constant over tim e.Conversely,the Ȗ changes over tim e because its m agnitude is proportional to the biom ass and is also affected by vegetation geom etry e.g., [12,13] M oreover, the Ȗ is know n to depend on the sensing configuration (e.g., frequency, view angle and polarization)e.g.,[14-16] M ethods forestim ating the Ȗ use eitherm ultiple channelm icrow ave data orancillary data.A direct estim ation ofthe Ȗ from m icrow ave data is preferred because the ancillary data needed ata globalscale forsoilm oisture retrievalm ay notbe available.H ow ever,its dependence on the instrum entparam eters com plicates the inversion of Ȗ from TB ¶V PHDVXUHG DW GLIIHUHQW IUHTXHQFLHV YLHZ DQJOHV DQG polarizations Large scale soil m oisture m onitoring studies e.g., [17-19] have, therefore, frequently adopted the ancillary data approach to determ ine the Ȗ,w hich has been extensively described in the scientific literature e.g., [20,21] For this characterization, the Ȗ is related to the vegetation optical depth (IJ), w hich is estim ated as a function of the vegetation w ater content (W ) and a crop-specific em piricalparam eter,b,w hich depends on the instrum entparam eters V arious im plem entations of this approach w ithin soil m oisture retrieval algorithm s have been reported.Forexam ple,Jackson etal.[8]used a land coverm ap to define foreach crop type a specific b value and utilized the N orm alized D ifference V egetation Index (N D V I)to estim ate the W Sim ilarly, Bindlish et al [22] adopted the N D V I as a proxy for the W , but inverted the b values from dual-polarized X -band (10.65 G H z)TB by assum ing thatthe IJ is the sam e forboth horizontally (H )and vertically (V )polarized data.This polarization dependence is taken into accountby the SM O S level2 processor [12] as its effect is expected to be m ore significant at the low er L-band frequency In Sensors 2010,10 6982 addition,a m ore sophisticated approach for m odeling the view angle dependence of IJ is included in SM O S processor because its TB ¶V DUH FROOHFWHG IURP GLIIHUHQW DQJOHV >@ )XUWKHU DSDUW IURP WKH N D V Ialso the LeafA rea Index (LA I)has been found to be a good estim atorforW [23]and is used for the SM O S soilm oisture retrieval A lthough iterative procedures for inverting the IJ have been developed e.g.,[11,22],initial values and uncertainty ranges for the IJ are still needed as input The selection of the appropriate param eterization for a specific land cover relies, how ever,often on param eter sets derived from TB m easurem ents collected during pastintensive field cam paigns e.g.,[16,20].By default,the validity of those param eterizations is restricted to the conditions for w hich they have been derived.M any of the pastfield cam paigns covered,forexam ple,a partofthe grow th cycle ofagriculturalcrops.Therefore, the tem poralevolution ofthe Ȗ and b param eterthroughoutthe grow th cycle is notfully understood This paper contributes to this understanding by analyzing the L-band H -polarized TB ¶V PHDVXUHG throughout the com plete 2002 corn (Zea m ays L.) grow th cycle The utilized data set has been collected atone of the fields of the Beltsville A griculturalResearch Center (BA RC) by an autom ated tow erm ounted L-band (1.4 G H z)radiom eter(called LRA D )starting from M ay 22nd tillthe beginning of Septem ber These radiom eter m easurem ents w ere supported by a detailed land surface characterization, w hich took place about once every w eek and included m easurem ents of the vegetation biom ass, soil m oisture and soil tem perature D espite m echanical difficulties w ith the scanning system ofLRA D thatproduced gaps in the data record,a totaloften days distributed overthe grow ing season of both radiom eter and ground m easurem ents are available covering a W range from 0.0 to 4.3 kg m í2 The objective ofthis investigation is to evaluate the variations in the Ȗ and the em piricalparam eterb over the m onitored corn grow th cycle First, the im pact of the surface roughness on the surface em ission is quantified using the LRA D TB ¶VRYHUEDUHVRLOFRQGLWLRQVDQGDQROGHUGDWDVHWFROOHFWHGDW the BA RC facility Subsequently, the Ȗ (and b param eter) are inverted from individual TB m easurem ents using the estim ated roughness param eterization, and m easured soil m oisture and soil tem perature In addition, an analysis is presented of the sensitivity of the derived b param eters for uncertainties in the LRA D TB and the assigned single scattering albedo (Ȧ) 2.TheoreticalBackground The starting point for the com putation of m icrow ave em ission from vegetated surfaces is the sem i-em piricalradiative transferapproach by M o etal.[24],w hich is based on the assum ption thatat L-band absorption is dom inantoverscattering, TBp 1 R J 1 J 1 Z T 1 R J T p s p p p v p s p s (1) w here,TBp is the polarized brightness tem perature, Rsp is the soilsurface reflectivity (= í em issivity), Ȗp is the transm issivity of vegetation, Ȧp is the single scattering albedo, Ts and Tv are the soil and canopy tem peratures,respectively,and superscriptand subscriptp indicates the polarization The firstterm on the righthand side of Equation (1)represents the m icrow ave em ission directly by vegetation and the radiation em itted by the vegetation reflected by the soil surface back tow ards the Sensors 2010,10 6983 sensor The second term quantifies the em ission contribution from the soil, corrected for the attenuation by the vegetation layer The solution to the radiative transfer equation requires param eterization of the vegetation and soil surface layerradiative transferproperties.Further,tem peratures ofthe vegetation and soilsurface layer are required.H ow ever,w hen assum ing the vegetation and soilsurface are in therm alequilibrium w ith each other,Ts and Tv can be considered equal;this condition occurs typically neardaw n.The required tem perature is then considered representative forthe em itting layer 2.1.Em ission from Soil The surface em issivity is typically described in term s ofthe surface reflectivity.This is convenient because the m icrow ave reflectivity under sm ooth surface conditions can be calculated using the Fresnelform ulas forreflectivity (R p ),w hich read forthe H and V polarization, R H T cosT H r sin2 T cosT H r sin T 2 RV T and H r cosT H r sin2 T H r cosT H r sin2 T 2 (2) w here,İr is the dielectric constantofsoil,ș is the incidence angle In this study,the approach described by W ang and Choudhury [25]has been adopted to accountfor the effectof surface roughness on the reflectivity.This approach involves tw o param eters,w here one param eter has an attenuating effect on the surface reflectivity and the other accounts for the depolarizing effectofthe surface roughness, Rsp T êơ 1 Q R p QR q º¼ exp h w ith h h0 G T (3) w here,h0 is roughness param eter given by 4k2ı2 w ith k DVWKHZDYHQXPEHU ʌȜ) and ı as the root m ean square (rm s) heightof the surface heightvariations,Q is a polarization m ixing factor,G (ș) is a function accounting for the angular dependence of surface roughness effecton surface em ission and superscriptq represents the polarization orthogonalto polarization p,w hich can be eitherH orV O riginally, W ang and Choudhury [25] took the function G (ș) equal to cos2ș H ow ever, W ang et al [26] have found that the dependence of cos2ș is m uch too strong and replaced it by G (ș)= 1.0 forbestfitting theirdata.The latteris initially adopted here 2.2.Vegetation Effects on SoilSurface Em ission W ithin the radiative transfer approach, vegetation effects are characterized by tw o param eters: transm issivity (Ȗ) and single scattering albedo (Ȧ).The Ȧ is a m easure for the fraction of attenuated radiation scattered from the canopy, Zp N sp N sp N ap (4) w here,N sp and N ap are the scattering and absorption coefficients,respectively These scattering and absorption coefficients can be obtained through application of the discrete m edium approach e.g., [27,28], in w hich individual com ponents of the vegetation layer (leaves and Sensors 2010,10 6984 stem s) are represented by elliptical and/or cylindrical dielectric scatterers In som e cases, the Ȧ is assum ed to be negligible or a variable dependenton the grow th stage,w hich can be determ ined from controlled experim ents w here all other variables (e.g., soil m oisture, tem perature of em itting layer, surface roughness and transm issivity)are m easured The transm issivity describes the am ountofsoilem ission passing through the vegetation layerand is an im portant variable for quantification of the effect of vegetation on m icrow ave em ission The one-w ay transm issivity through the canopy layeris form ulated as, Jp Đ W p à expă © cosT ¹ (5) w here,IJp is the polarization dependentopticaldepth orcanopy opacity,w hich can be calculated using, Wp kep H v (5) no Im fpp (6) w ith kep 4S O w here,H v is the canopy height,kep is a polarization dependentextinction coefficient,no is the num ber of phytoelem ents per unit volum e, Ȝ is the w avelength and Im fpp is the im aginary part of the polarization dependentscattering m atrix ofthe phytoelem ents in the forw ard direction Severalstudies [15,16,20]have show n thatIJp can be related to the vegetation w atercontentas, W p bp W (7) w here,W is the vegetation w ater content(kg m í2) and bp is an em piricalparam eter varying w ith crop type,canopy structure,w avelength,and polarization Equation (7) requires inform ation aboutthe W and bp param eters for differenttypes of vegetation This approach has been frequently used forsoilm oisture retrievalpurposes e.g.,[1,8,19]and has been proposed as part of the soil m oisture retrieval algorithm s for current and future m icrow ave radiom eters e.g.,[29].The SM O S level2 soilm oisture retrievalprocessor adopts a sim ilar approach relating the IJp to the leafarea index (LA I)instead ofthe W [30] 3.The O PE Experim ent 3.1.Site D escription The presentstudy w as conducted atO ptim izing Production Inputs forEconom ic and Environm ental Enhancem ent (O PE3) test site m anaged by the U SD A -A RS (U nited States D epartm ent of A griculture-A gricultural Research Service) [31] The site consists of four adjacent w atersheds w ith sim ilar surface and sub-surface soil and w ater flow characteristics and covers an area of 25 near Beltsville,M aryland (Figure 1).Each ofthe fourw atersheds is form ed from sandy fluvialdeposits and has a varying slope ranging from 1% to 4% The soiltexturalproperties are classified as sandy loam w ith 23.5% silt,60.3% sand,16.1% clay,and bulk density of1.25 g cm í3.A detailed description ofthe research activities can be found athttp://hydrolab.arsusda.gov/ope3.(V erified D ecem ber23,2009) Sensors 2010,10 6985 Figure 1.Location and schem atization ofthe O PE3 rem ote sensing experim entalsetup in 2002 O PE study area Experim entalSetup N W ashington D C N 3.2.G round M easurem ents The in-situ m easurem ent strategy w as designed to provide ground inform ation to supplem ent the radar and radiom eter data acquisitions,and took place every W ednesday,rainy days excluded.In this paper,an analysis of the radiom eter observations is presented.A description of the radar data set is given in Joseph etal.[32] D uring the field cam paign (M ay 10 to O ctober 2, 2002) representative soil m oisture, soil tem perature,vegetation biom ass (w etand dry)and surface roughness m easurem ents w ere taken around the radiom eter footprints Soil m oisture and soil tem perature m easurem ents w ere collected at tw enty-one sites located at the edge of a 67.1 m u 33.5 m rectangular area depicted in Figure V egetation biom ass and surface roughness m easurem ents w ere taken around the study area at representative locations Soilm oisture and soiltem perature Soil m oisture w as m easured using gravim etric, portable im pedance probe²D elta-T theta probe (The U S G overnm ent does not endorse any specific brand of im pedance probe for m easuring soil m oisture or any specific brand of digitaltherm om eters),and buried im pedance probe (Tim e D om ain Reflectom etry (TD R)) techniques Soil sam ples of the top 6-cm soil layer w ere collected at the beginning of each day in conjunction w ith the theta probe m easurem ents prim arily for calibration purposes.Theta probe m easurem ents w ere collected typically at 8:00,10:00,12:00 and 14:00 hours Sensors 2010,10 6986 (U SA Eastern) The buried TD R probes w ere installed at locations R5,R11 and R18 (Figure 1) at various depths (5,10 and 20 cm )and insertion angles (horizontal,vertical,and 45 degrees) The soil dielectric constant (İr) m easured by the theta probe w as converted to volum etric soil m oisture (M v) values by fitting a linear regression function through the follow ing relationship (Figure 2a), Hr a0 a1 M (8) v w here,a0 and a1 are regression param eters Figure (a) Com parison of the calibrated theta probe against the gravim etric M v; (b) Com parison of the theta probe m easured İr against the calculations m ade using the D osbon soil m ixing m odel; (c) M v m easured by the theta probe, TD R and gravim etric sam pling technique plotted againsttim e 25 (a) (b) 20 Dobson diel const 0.2 0.1 15 10 Data points 1:1 line 0 0.1 0.2 Gravimetric Mv [m3 m-3] 0.3 10 15 0.4 30 0.3 Mv [m3 m-3] 20 25 Theta probe diel const 60 Rainfall [mm d-1] Theta probe Mv [m3 m-3] 0.3 0.2 Gravimetric Mv 0.1 Theta Probe Mv TDR Mv (c) 0.0 5/1/02 Rainfall 6/1/02 7/1/02 8/1/02 Date [mm/dd/yy] 9/1/02 10/1/02 W hile general soil texture-specific param eters are available [33], a site specific calibration w as perform ed To achieve this, soil m oisture determ ined gravim etrically from the soil sam ples w as converted to M v and used w ith concurrent probe observations to fit for each site a set of a0 and a1 values.Com parison of the calibrated theta probe M v values w ith the gravim etric M v (see Figure 2a) gives a rootm ean squared error (RSM E) of 0.024 m m í3,w hich is com parable to calibration errors Sensors 2010,10 6987 obtained w ith theta probe observations collected in severalrem ote sensing cam paigns [34].In addition, Figure 2b show s the İr m easured by the Theta probe plotted againstthe İr calculated w ith soilm ixing m odel of D obson et al [35] using the soil texture and the gravim etric M v.The RM SE of 1.87 and coefficientof determ ination (R 2) of 0.77 com puted betw een the m easured and calculated İr indicates that both m ethods for quantifying İr are in agreem ent w ith each other Further, the M v determ ined using the gravim etric,Theta probe and TD R probe techniques are displayed as tim e series in Figure 2c forcom parison purposes.A s show n by the plot,sim ilartem poralsoilm oisture variations are observed by the three m easurem entapproaches,w hich justify the use ofeach oftheirproducts 0.3 30 (a) mean Mv 60 Mv stdev Rainfall 0.1 Stdev [m3 m-3] 0.2 0.1 0 Mean Soil Temp [oC] 5/1/02 30 6/1/02 7/1/02 8/1/02 9/1/02 10/1/02 (b) 20 mean cm Tsoil mean cm Tsoil cm Tsoil stdev 10 cm Tsoil stdev 5/1/02 Stdev [oC] Mean Mv [m3 m-3] 0.4 Rain [mm d-1] Figure 3.M ean and standard deviation oftw enty-one soilm oisture (a)and,3-cm and 7-cm soiltem perature;(b)m easurem ents collected around the radiom eterfootprints 6/1/02 7/1/02 8/1/02 Date [mm/dd/yy] 9/1/02 10/1/02 Soiltem perature m easurem ents w ere taken m anually atsoildepths of 3- and 7-cm ateach of the tw enty-one sam pling locations (annotated as R1 to R21 in Figure 1) throughoutthe experim entusing Extech Instrum ents digitalstem therm om eters.O n intensive sam pling days the soiltem peratures w ere m easured at 8:00, 10:00, 12:00, 14:00 hours, and the m easurem ents on other days w ere taken approxim ately every tw o days at8:00 and 14:00 hours A lthough the study area w as selected to m inim ize the effects of land surface heterogeneity,sm all surface height and soil texture variations could potentially influence the representativeness of the m easured soil m oisture and tem perature for the radiom eter footprints These effects are studied by presenting the tem poral evolution of the m ean and standard deviation (stdev) of the tw enty soil m oisture and soiltem perature m easurem ents in Figure 3.Figure 3a show s thatthe m ean soilm oisture changes in response to antecedent rain events A lso, the soil m oisture stdev varies over tim e Sensors 2010,10 6988 from 0.003 m m í3 under extrem e dry conditions to 0.036 m m í3 in the m id soilm oisture range.O n average,how ever,the stdev rem ains quite stable around values ofabout0.020 to 0.030 m m í3,w hich is com patible w ith the Theta probe calibration uncertainty of 0.024 m m í3 Further, the spatial tem perature variability at soil depths of cm (Figure 3b) and cm (Figure 3c) is quite low w ith averaged stdev values of0.73 and 0.58 oC,respectively.G iven the fairly stable soilm oisture stdev and low tem perature stdev observed,the spatialheterogeneity around the footprintis expected to have only a m inoreffecton the representative m ean ofthe tw enty-one m easurem ents forthe radiom eterfootprint The m ean soilm oisture and soiltem perature values are,therefore,used forfurtheranalysis V egetation Corn w as planted on A pril 17, reached peak biom ass around July 24 and w as harvested on O ctober V egetation biom ass and m orphology w ere quantified through destructive m easurem ents applied to m area (approxim ately 12 plants) once every w eek at8:00 am The w ater content,fresh and dry biom asses w ere determ ined separately for the individual plant constituents, such as leaves stem s and cobs (w hen present) Figure 4a show s the developm ent biom asses and w ater content of the total plant over tim e and Figure 4b illustrates the tem poral evolution of the w ater content in individual plant com ponents It follow s from Figure 4b that in the beginning of the corn grow ing season,the canopy w as prim arily m ade up ofleaves and stalks.In the m iddle ofthe grow ing season the stem contribution becom es m ore GRPLQDQWDQGFREV¶ZDWHUFRQWHQWLQFUHDVHVWROHYHOVH[FHHGLQJWKHOHDIFRQWULEXWLRQ1HDUVHQHVFHQFH w atercontentin the leaves is reduced further,w hereas the contribution ofthe cobs to the totalbiom ass rem ained constant Figure (a) Total plant w ater content, fresh and dry biom ass plotted against tim e (b) W atercontentin the leaves,stem s and cobs plotted againsttim e.The m arkers indicate the dates atw hich m easurem ents w ere m ade 6.0 Fresh biomass Dry biomass Water content 6.0 (a) Water content [kg m-2] Biomass [kg m-2] 8.0 4.0 2.0 0.0 5/1/02 4.0 Total plant Leaves Stems Cobs (b) 2.0 0.0 6/1/02 7/1/02 8/1/02 9/1/02 10/1/02 Date [mm/dd/yy] 5/1/02 6/1/02 7/1/02 8/1/02 9/1/02 10/1/02 Date [mm/dd/yy] Surface roughness D uring the experim ent surface roughness w as characterized on M ay 25 using the grid board technique.A 2-m eter long grid board w as placed in the soiland photographs w ere taken w ith the soil surface in front.In total,ten surface heightprofiles w ere recorded.The surface heightprofile in these pictures w as digitized at a 0.5-cm interval,from w hich tw o roughness param eters w ere derived: the Sensors 2010,10 6989 rm s of the surface height and the correlation length (L).The averaged rm s height and L for the ten observed surface roughness profiles w ere found to be 1.62 and 12.66 cm ,respectively.Figure show s an exam ple of a photograph taken for this roughness characterization and lists the roughness param eters calculated from the digitized surface heightprofiles Figure The left panel show s an exam ple of a picture taken for surface roughness characterization and the rightpanellists the derived surface roughness param eters Listofsurface roughness param eters derived from digitized surface heightprofiles rm sheight[cm ] L [cm ] Profile 1.11 5.18 Profile 0.81 6.35 Profile 0.95 6.39 Profile 0.75 3.22 Profile 0.74 4.52 Profile 2.35 8.95 Profile 2.46 12.20 Profile 1.95 10.75 Profile 1.91 8.16 Profile 10 1.68 6.84 3.3.Radiom eter The deployed radiom eterw as a dual-polarized L-band passive m icrow ave sensor,called LRA D The instrum entw as m ounted on a portable 18 m tow er and w as designed to collectdata autom atically (for this experim entevery hour) atfive incidence angles (25,35,45,55,and 60 degrees) and three azim uth angles over a range of 40 degrees.LRA D had a dB beam w idth of approxim ately 12 degrees,w hich corresponds to footprints varying from 4.5 to 15.5 m eters for the 25 to 60 degrees incidence angle range M echanical difficulties w ith the scanning system restricted the LRA D data collection, and produced considerable gaps in the season-long record N evertheless, ten days of com plete record (ground m easurem ents and radiom eterobservations)w ere available forthe presentanalysis Each LRA D data run consisted of a pre-calibration,a m easuring sequence,and a post-calibration D uring each ofthe tw o calibration periods one m icrow ave observation w as acquired from a m icrow ave absorbertargetofknow n tem perature (hottarget) and one m icrow ave observation w as acquired ofthe sky (cold target), w hich has at L band an TB of ~5 K (3 K cosm ic background radiation and K atm ospheric contribution) These tw o so-FDOOHG ³KRW´ DQG ³VN\´ WDUJHW REVHUYDWLRQV FDQ EH XVHG WR calibrate,through linearinterpolation,the radiom eterobservations ofthe land surface using, TBp Thot Tsky U hot U sky U p Tsky Thot Tsky U hot U sky U sky (9) w here TB is the brightness tem perature [K ],T indicates the tem perature [K ] of the specified targetand U represents the LRA D voltage observations [V olt] w ith subscripts hotand sky indicating the hotand sky targetproperties Sensors 2010,10 6990 For processing the LRA D m easurem ents to TB ¶V WKH SUH-calibration w as used, w hile the post-calibration w as only em ployed to detect anom alous values The estim ated uncertainty of the calibrated H -polarized TB is about ±1.0 K W hile m easurem ents w ere also collected for vertical polarization, there rem ain som e unresolved issues w ith respect to the calibration of these m easurem ents.Thus,V -polarized m easurem ents are notbeing presented atthis tim e 4.R esults 4.1.Surface Roughness Estim ation U sing H -Polarized TB W ithin the bare soil em ission m odelby W ang and Choudhury [25],surface roughness effects are characterized by:(1)m odification ofthe reflectance (h param eter),and (2)redistribution ofthe H -and V -polarized em itted radiation (Q param eter) Since only reliably calibrated H -polarized TB m easurem ents are available for analysis,the Q param eter is om itted i.e.,(Q = 0),w hich essentially reduces the em ission m odelto the one proposed by Choudhury etal.[36].This form ulation has been adopted previously in severalotherstudies i.e.,[17,22].Based on this assum ption,the h param etercan be estim ated from H -polarized TB ảVPHDVXUHGRYHUEDUHVRLOXVLQJ ê TBH ô1 ằ Ts ẳ êơ R H T ẳ exp h (10) w here, TBH is the H -polarized brightness tem perature, TS is the soil tem perature, R H is the H -polarized Fresnelreflectivity Table Surface param eters obtained through inversion of H -polarized TB observations acquired overbare soilconditions h = h0Âcos ș h = h0Âcos ș h = h0 h = h0ÂVHFș h = h0Âsec2 ș 35 degrees 0.641 0.525 0.429 0.352 0.288 View angle 45 degrees 0.867 0.613 0.434 0.307 0.217 60 degrees 1.663 0.832 0.416 0.208 0.104 The LRA D observations during the O PE3 cam paign started on M ay 22,w hen corn crops had just em erged and the totalfresh biom ass w as less than 0.04 kg m í2.Forthese low biom ass conditions,the m easured TB ¶Vare used to estim ate the h param eterw hereby the m ean oftw enty-one soiltem peratures m easured at a cm soil depth is adopted as Ts U nfortunately, for this part of the experim ent, the m icrow ave observations w ere only collected from view angles of 35, 45 and 60 degrees The h param eters inverted for these view angles are given in Table for G (ș) functions equal to cos2 ș, cos ș,1,sec ș and sec2 ș The derived h param eters fallw ithin the range thathas been reported previously.W ang etal.[26] reported a 0.00±0.53 h param eterrange forsurfaces w ith a rm s heightvarying from 0.21 to 2.55 cm for Sensors 2010,10 6991 a sim ilarsetting.Considering an averaged rm s heightof1.62 cm w as m easured around the radiom eter footprint,the h param etervalues obtained from the LRA D observations appears reasonable A pointofdiscussion could,how ever,be the angulardependence ofthe h param eter.This is absent for the 35 to 60 degrees view angle range,w hich is in agreem entw ith previous reports e.g.,[26,38] A n angulardependence is som etim es expected because w hen a radiom eterobserves the land surface at different angles surface roughness m ay have a different im pact on the surface em ission, w hile recognizing thatEquation (10)is also an approxim ation [30].H ow ever,the angulardependence ofthe h param etercould also be a resultfrom the assum ption Q = 0.The Fresnelreflectivities forthe H -and V -polarization are both a function of the incidence angle; excluding one of the tw o polarization com ponents, as is done by assum ing Q = in Equation (3), induces an angular dependence of the h param eter 4.2.Surface Roughness Param eter Estim ation Based on D ual-Polarized TB The surface roughness param eter h from the presentdata setdem onstrates an angular dependence that is equal to adopting G (ș) = (see Table 1) A lim itation of the present data set is that only H -polarized TB observations are available to som e degree ofconfidence.Forretrieving the h param eter from these TB values Q w as assum ed zero, w hich m ight alter the angular dependency as discussed above (m ixing of polarization) To elaborate on these findings, dual polarized L-band (~1.4 G H z) radiom eter data sets collected over bare soils w ithin the general area of the present study [23] are utilized to inverth and Q sim ultaneously The m ethodology used to retrieve the Q and h param eters has been adopted from W ang and Choudhury [25],w hich is based upon the follow ing tw o relationships, X T Y T 1 V TNB T TNBH T V 1 êơTNB T TNBH T ẳ V H êTNB T TNB T ẳ 2ơ ê R H T RV T º 2« H » 1 2Q V R T R T ẳ ê R H T RV T º¼ exp hG T 2¬ (11) (12) w here,TNBp is the norm alized brightness tem perature forpolarization p,according to TBp Ts ,X (ș)is the surface roughness coefficientfor deriving the Q param eter,< ș is the surface roughness coefficient forderiving the h param eter Equations (11)and (12)can be rew ritten to give the Q and h explicitly resulting in, Q ê X T ô1 ằ ơô êơ P T ẳ ẳằ and P T ê R H T RV T ô H ằ V R T R T ẳ (13) w ith h ê 2Y T ln ô H ằ V ôơ êơ R T R T º¼ »¼ G T (14) The data set described in W ang et al [26] includes ground m easurem ents of soil m oisture and tem perature observed at various depths: 0±0.5,2.5±5.0,5.0±10.0 cm for soil m oisture and 1.25,2.5, Sensors 2010,10 6992 7.5 and 15.0 cm for soil tem perature In addition, dual-polarized TB observations w ere collected at view angles of 10,20,30,40,50,60 and 70 degrees.These m easurem ents have been collected over soil surfaces w ith different roughness characteristics For this investigation, a sm ooth and a rough surface are included in the analysis w ith a m easured rm s height of 0.73 and 2.45 cm , respectively Because the presentdata setincludes radiom eter observations for an incidence angle range betw een 35 and 60 degrees,only the TB m easured overthe 20 to 60 degrees incidence angle range are utilized Figure h-param eter as a function of incidence angle calculated from dual-polarized L-band TB ¶V PHDVXUHG RYHU (a) sm ooth bare soil surface and (b) rough bare soil surface (c)Q -param eters as a function ofthe incidence angle forsam e sm ooth and rough surfaces h-parameter [-] 0.4 (a) 1.6 0.3 1.2 0.2 0.8 0.4 0.1 cos2(theta) cos(theta) G(theta)=1 cos2(theta) cos(theta) G(theta)=1 (b) 0.0 20 40 Incidence angle [degrees] 20 60 40 Incidence angle [degrees] 60 (c) Q-parameter [-] 0.4 0.3 0.2 0.1 Rough Smooth 0.0 20 40 Incidence angle [degrees] 60 The extensiveness of the radiom eter and ground m easurem ents perm its all unknow ns in equations (13) and (14) to be derived,and allow s the com putation of surface roughness param eters Q and h In analogy w ith the previous roughness com putations, the soil m oisture content integrated over 0±5.0 cm has been used to com pute the relative dielectric constant and the soil tem perature at2.5 cm has been used to derive the norm alized brightness tem perature.The resulting h param eters are plotted as a function ofthe incidence angle forthe rough and sm ooth bare soilsurface in Figures 7a and 7b respectively,w hereas the com puted Q values are show n as a function ofthe incidence angle for both the rough and sm ooth surface in Figure 7c.The h-param eters show n in Figure 7a and 7b have been com puted assum ing three differentG (ș)relationships,w hich are:G (ș)= cos2ș,cos ș and 1.0 Figures 7a and 7b show a differentangular behavior of the em ission m easured over the rough and the sm ooth surface For the rough surface, it is observed that the function G (ș) = cos ș results in angular independent h param eter H ow ever, none G (ș) functions are able to suppress the angular Sensors 2010,10 6993 dependence of the h param eter from the sm ooth surface, w hile G (ș) = cos2ș provides the best approxim ation.Further,an angulardependency ofQ param eteris noted in Figure 7c forboth the rough and sm ooth surfaces The discussion above and previous results e.g.,[26,36,38]indicate thatconsistencies in the angular dependence ofroughness effecton m icrow ave em ission are difficultto identify.H ence,forSM O S soil m oisture processorh is approxim ated by, h h0 cosN RP T (15) w here, N RP quantifies the angular dependence of h0, w hich is also assum ed to be polarization dependent The param eters,h0 and N RH ,have been fitted to m atch ourm ulti-angular data collected overnearly bare soilconditions.The obtained param eter values,and RM SE com puted betw een the m easured and retrieved TB ¶VDQGM v¶V 506(Tb and RM SD m v)are presented in Table 2.In addition,the optim ized h0 as w ellas the RM SETb and RM SD m v obtained w ith the m ore frequently used N RP values are given in Table Table 2.h0 param eter inverted using m ultiangular H -polarized TB ¶V m easured over bare soil and assum ing different N RH values, in bold are the h0 and N RH param eters sim ultaneously inverted from the m ultiangulardata RM SE N RH N R H = 0.05 (opt.)* N RH = í2 N RH = í1 N RH = N RH = N RH = * h0 0.411 0.104 0.277 0.407 0.613 0.784 TB [K ] 1.205 11.225 6.155 1.208 7.441 13.734 M v [m m í3] 0.0053 0.0616 0.0074 0.0065 0.0082 0.0074 N RH and h0 are calibrated sim ultaneously A n analysis ofthe param etervalues show n in Table dem onstrates the advantage ofincorporating the N RP param eter.506(¶V EHWZHHQ WKH PHDVXUHG DQG VLPXODWHG TB ¶V LQFUHDVH IURP DERXW WR m ore than 10.0 K w hen the N RP is changed from to í2 or +2 Surprisingly, this reduction in the ability to sim ulate TB ¶V RQO\ UHGXFHV WKH VRLO PRLVWXUH UHWULHYDO DFFXUDF\ significantly w hen N RP is taken equalto í2.This is explained by the factthatthe M v is retrieved by using as costfunction the RM SE com puted betw een the TB sim ulated and m easured from differentview angles ata given tim e step.For less negative and positive N RP values,the underestim ation of m easured TB atlow (or high) view angles is com pensated by an overestim ation athigh (orlow ) view angles.H ence,the increase in the retrievaluncertainty is for various N RH values m uch sm aller than w ould be expected based on the PRGHO¶VDELOLW\WRVLPXODWH7B ¶VItshould,how ever,be noted thatin this case only bare soilconditions are considered and,thus,results m ay be differentundervegetation conditions Sensors 2010,10 6994 4.3.Estim ation ofthe H -polarized Transm issivity W hen soil m oisture and surface tem perature are know n, H -polarized transm issivity (Ȗh) can be estim ated through the inversion of Equation (1) assum ing that tem poral variations in the roughness param eters are sm alland the Ȧ equalto zero.Estim ates ofthe Ȗh are only presented forretrievals from H -polarized TB ¶VPHDVXUHGLQWKHHDUO\PRUQLQJ(around 8:00 A M )because atthattim e ofthe day the soilsurface and vegetation are typically found to be in therm alequilibrium e.g.,[40].This assum ption perm its using a single so-FDOOHG àeffectiveả tem perature as input for Equation (1), for w hich the tem perature m easured at a cm soil depth is adopted Further, for calculation of the Fresnel reflectivity,the İr LVREWDLQHGWKURXJKDSSOLFDWLRQRI'REVRQ¶VVRLOPL[LQJPRGHO>5]w ith inputofsoil texturalproperties and the m easured soilm oisture Table H -polarized transm issivities and b param eters estim ated over the 2002 corn grow th cycle using m ultiangularbrightness tem peratures D ate M ay 29,2002 June 5,2002 June 19,2002 June 26,2002 July 3,2002 July 9,2002 July 12,2002 August21,2002 August30,2002 W kg m í2 0.1 0.3 1.9 3.1 3.7 4.2 4.3 2.6 2.0 Transm issivity 35o 45o 60o 0.945 0.951 0.967 0.857 0.878 0.881 0.784 0.830 0.763 0.678 0.684 0.641 0.695 0.679 0.629 0.640 0.552 0.556 0.639 0.532 0.517 0.783 0.758 0.736 0.821 0.786 0.732 35o 0.431 0.423 0.105 0.101 0.081 0.088 0.085 0.078 0.081 B param eter 45o 60o 0.356 0.167 0.306 0.211 0.070 0.071 0.085 0.070 0.075 0.063 0.101 0.070 0.103 0.076 0.076 0.060 0.086 0.079 The retrieved Ȗh foreach day and view angles of35,45 and 60 degrees are given in Table and are plotted in Figure 8a againstthe totalplantW In Figure 8a,Ȗh com putations are also presented for an assum ed b param etervalue of0.117 m kgí1,w hich is the m edian ofL-band corn b values presented in Jackson and Schm ugge [20].Further,b param eters have been derived from the retrieved Ȗh¶VZKLFK are given in Table and plotted againstthe totalplantW in Figure 8b.Itshould be noted thatm ostb param eters have previously been derived fordense corn canopies w ith W in the range 1.2í6.0 kg m í2 A com parison of b param eters derived for M ay 29 and June (W = 0.1 and 0.3 kg m í2) against previously reported values is,therefore,not optim al.The field conditions observed from June 19 to A ugust 30 (W = 1.9 ± 4.3 kg m í2) are, how ever, com patible in term s of biom ass to corn canopies referred to in these previous investigations Sensors 2010,10 6995 Figure H -polarized corn transm issivities (a) and b param eters (b) inverted from TB ¶V m easured atincidence angles of35,45 and 60 degrees plotted againstthe totalplantW 0.5 (a) b parameter [m2 kg-1] Transmissivity [-] 0.8 0.6 0.4 35 degrees 45 degrees 60 degrees Theory 35 degrees Theory 45 degrees Theory 60 degrees 0.2 (b) 0.4 0.3 0.2 0.1 0 W [kg m-2] W [kg m-2] Figures 8a show s that the retrieved Ȗh follow s a different pattern than expected based on the literature.A tthe beginning of the grow th cycle,the Ȗh is sm aller than expected,w hile closer to peak biom ass the Ȗh is larger.In term s of the b param eter,the obtained values are higher than the literature reports justafter em ergence of the corn crops and som ew hatlow er athigher W levels (>1.9 kg m í2) The dependence ofthe b param eteron W can be argued based on previous investigations.Le V ine and K aram [41],am ong others,have show n that the attenuation by canopies com posed of elem ents w ith sim ilar dim ensions as the w avelength is also specific to the vegetation m orphology A s changes in biom ass (or W ) are typically associated w ith differentgrow th stages and also architecturalchanges in the canopy,the b param etercan be expected to vary throughoutthe grow th cycle Table 4.H -polarized transm issivities and b param eters inverted from TB ¶VPHDVXUHGXQGHU sparsely (M ay 29th, W = 0.1 kg m í2) and densely (July 9th, W = 4.2 kg m í2) vegetated conditions and perturbed by +1.0,0.0 and í1.0 K ,respectively M ay 29th W = 0.1 kg m í2 July 9th W = 4.2 kg m í2 Transm issivity 45o 60o 35o 0.958 0.958 0.973 0.355 0.300 0.139 TB 0.949 0.951 0.967 0.431 0.356 0.167 TB + 1.0 K 0.940 0.943 0.962 0.510 0.413 0.195 TB í 1.0 K 0.690 0.593 0.577 0.073 0.089 0.066 TB 0.640 0.552 0.556 0.088 0.101 0.070 TB + 1.0 K 0.584 0.506 0.535 0.106 0.116 0.075 TB í 1.0 K 35o b-param eter 45o 60o O n the other hand,it should be noted that the presented Ȗh¶VDQG b param eters are also subjectto various sources of uncertainty em bedded w ithin the inversion procedure.For exam ple,corn crops at the beginning of the grow ing season are very sm all,w hich lead to relatively large uncertainties in the m easured W M oreover,the contribution ofthe vegetation em ission to the m easured TB is also sm allat the early grow th stage.U ncertainties in the m easured TB m ay have,therefore,a large im pact on the Sensors 2010,10 6996 derived b param eters To dem onstrate the im pact of such TB uncertainties on the derivation of b param eters from m easurem ents acquired oversparse and dense vegetation,the Ȗh¶VKDYHEHHQLQYHUWHG after perturbing the TB m easured on M ay 29th (W = 0.1 kg m í2) and July 9th (W = 4.2 kg m í2) by ±1.0 K The obtained Ȗh¶VDQGb param eters for these tw o dates are given in Table 4.These results confirm thatundersparsely vegetated conditions TB uncertainties have a largerim pacton the derived b param eters than under densely vegetation conditions The b values retrieved for M ay 29th range from 0.355 to 0.510 m kgí1 for the 35 degrees view angle,w hile for the sam e angle the b param eter from July 9th range from 0.073 to 0.106 m kgí1 The som ew hat higher Ȗh¶V DQG ORZHU b param eters) obtained over m ore dense vegetation are explained by the effects ofscattering w ithin the canopy,w hich has notbeen considered as Ȧ = 0.0 has been assum ed.W hen the attenuation by vegetation is sm all,the scattering w ithin the canopy can be assum ed negligible because the em ission by vegetation is sm alle.g.,[14].This justifies using Ȧ = 0.0 A s the biom ass increases, vegetation em ission also increases and scattering w ithin the canopy w ill have a m ore im portantim pacton the m easured TB The previously reported Ȧ values tabulated in V an de G riend and W igneron [37]m ay reach forL-band and corn up to values of0.13 Table 5.Single scattering albedo (Ȧ) inverted from LRA D TB m easured on June 9,2002 (W = 4.2 kg m í2)assum ing a range b param eters from 0.10 to 0.15 m kgí1 b-param eter m kgí1 0.10 0.11 0.12 0.13 0.14 0.15 Single albedo 35o 0.014 0.016 0.018 0.020 0.021 0.022 scattering 45o 0.017 0.021 0.024 0.027 0.028 0.030 60o 0.033 0.037 0.040 0.043 0.044 0.045 To quantify this effectofscattering underdensely vegetated conditions,the Ȧ is inverted,instead of Ȗh,from the TB ¶V m easured on July 9th for assum ed b param eters of 0.10, 0.11, 0.12, 0.13, 0.14 and 0.15 m kgí1.The obtained Ȧ¶V DUH JLYHQ LQ 7DEOH , w hich illustrate the num erical correlation betw een the param eters b and Ȧ w ithin TB sim ulations using Equation (1); nam ely for sm all b param eters,Ȧ is also sm all.Further,itis noted thatthe inverted Ȧ¶VDUHGHSHQGHQWRQWKHYLHZDQJOH This can be argued for since scattering w ithin the canopy is affected by orientation of scatterers (e.g.,stem s,leaves and cobs)relative to the view angle e.g.,[27,28] The previous discussion on the effects of vegetation on L-band H -polarized TB ¶Vdem onstrates that the strength ofscattering and absorption w ithin a corn canopy changes overthe grow th cycle.This can be attributed to cKDQJHVLQ WKHFDQRS\¶s architecture as the corn crops develop.A s a result,the IJh is found to be a nonlinear function of the W , w hile m ost of current soil m oisture retrieval algorithm s adopt a linear relationships.To evaluate how this assum ption influences the reliability of retrievals, soilm oisture is inverted by m inim izing RM SE betw een sim ulated and m easured TB ¶VIRUYLHZDQJOHV Sensors 2010,10 6997 of 35, 45, and 60 degrees and assum ing a constant b and Ȧ of 0.12 m kgí1 and 0.0 based on [30],respectively Figure Soil m oisture m easurem ents and retrievals obtained by assum ing a constant b param eterand Ȧ of0.12 m kgí1 and 0.0,respectively RMSE = 0.054 m3 m-3 R2 = 0.824 0.3 0.2 0.1 W [kg m-2] Soil moisture [m3 m-3] 0.4 0 6/1/02 7/1/02 8/1/02 9/1/02 Measured Retrieved W The retrieved and m easured soil m oisture is plotted against tim e along w ith the total plant W in Figure The plot show s an underestim ation of m easured soil m oisture over sparse vegetation (W < 1.0 kg m í2) and an overestim ation for densely vegetated conditions (W > 1.5 kg m í2).A s the contribution of vegetation on both TB m easurem ents and sim ulations is m ore significant at a high biom ass, the im perfect vegetation param eterization leads to a larger overestim ation for dense vegetation as com pared to the underestim ation found forsparse vegetation;RM SE = 0.021 m m í3 for W < 1.0 kg m í2 and RM SE = 0.065 m m í3 for W > 1.0 kg m í2.Based on these results it m ay be concluded that consideratioQ RI WKH FDQRS\¶V DUFKLWHFWXUH IRU GHWHUPLQLQJ WKH YHJHWDWLRQ SDUDPHWHUV w ill assist in further im proving the reliability of soil m oisture retrievals especially over dense vegetation 5.C oncluding R em arks In this investigation, the H -polarized TB ¶V PHDsured by a tow er m ounted L-band (1.4 G H z) radiom eter w ere used to analyze the vegetation effects on surface em ission throughoutthe 2002 corn grow th cycle.Concurrentw ith the radiom eterm easurem ents an extensive land surface characterization took place about once a w eek including soil m oisture, soil tem perature and vegetation biom ass m easurem ents.O verthe period from M ay 22 to A ugust30,ten days w ith a com plete record ofground and radiom eterm easurem ents are available foranalysis covering a vegetation w atercontent(W )range of0.0 to 4.3 kg m í2 The roughness param eter, h,needed to correct for the effects of surface roughness, w as inverted from H -polarized TB m easured early in the corn grow ing season over essentially a bare soil surface Since V -polarized TB m easurem ents w ere not available for this investigation, the surface em ission Sensors 2010,10 6998 m odelby Choudhury etal.[34] (assum ing Q = 0.0) w as adopted and different G (ș) functions w ere used to analyze the angular dependence of h W hile recognizing that both V - and H -polarized reflectivities depend on the view angle,the assum ption Q = 0.0 could affected the obtained angular dependency of h.Therefore,a dual-polarized L-band radiom eter data setfrom 1981 [26] w as used to investigate the im pact of assum ing Q equal to 0.0 It w as found that even w ithin this com plete radiom eterdata sets consistencies in the angulardependence ofthe h are difficultto identify,w hich is in line w ith the param eterization G (ș) = cosN RP(ș) adopted for the SM O S level soil m oisture processor U sing this form ulation, a good agreem ent w as obtained betw een the m easured and com puted TB Based on the derived surface roughness form ulation, H -polarized corn transm issivities (Ȗh) have been retrieved using the radiative transfer equation and assum ing the single scattering albedo (Ȧh) equalto zero.The derived Ȗh¶VZHUHFRQYHUWHGLQWRb param eter values using the m easured totalplant W For sparse vegetation,the obtained IJh¶VDQGb param eters w ere found to be larger than the values reported in the literature This is partly explained by the fact that under low biom ass conditions TB uncertainties resultin a particularly large uncertainty in the derived b param eter.Fordense vegetation, the inverted b param eters w ere som ew hat sm aller than expected, w hich w as attributed to scattering w ithin the canopy thatw as notaccounted for,since Ȧ w as initially assum ed to be zero.By assum ing thatthe corn b param etervaries from 0.10 to 0.15 m kgí1,the Ȧh w as derived from TB m easurem ents For this range of b param eters,the obtained range in Ȧh¶Vis in agreem entw ith literature reports,but displays a strong angulardependence This study show s that the strength of scattering and absorption w ithin a corn canopy changes throughoutthe grow th cycle,w hich can be largely attributed to changes in architecture of vegetation layer.Forfurtherim provem entofthe soilm oisture retrievalreliability overdense vegetated conditions WKH FDQRS\¶V DUFKLWHFWXUH VKRXOG EH WDNHQ LQWR FRQVLGHUDWLRQ IRU GHWHUPLQLQJ YHJetation param eters A nalysis of additionalradiom eter data sets and sim ulations by advanced vegetation scattering m odels is recom m ended to obtain a m ore thorough understanding of the behavior of the b param eters throughoutthe grow th cycle A cknow ledgem ents The authors w ould like to acknow ledge thatthe field cam paign w as financially supported through N A SA and w e w ould like to thank various students forparticipating in the field cam paign R eferences and N otes Jackson, T.J M easuring large scale surface soil m oisture using passive m icrow ave rem ote sensing.H ydrol.Process.1993,7,139-152 W igneron,J.P.;K err,Y ;W aldteufel,P.L-band m icrow ave em ission of the biosphere (L-M EB) m odel:D escription and calibration againstexperim entaldata sets over crop fields.Rem ote.Sens Environ.2007,107,639-655 O w e,M ;D e Jeu,R.A M ;H olm es,T.R.H M ulti-sensorhistoricalclim atology ofsatellite-derived globalland surface m oisture.J.G eophys.Res.2008,113,doi:1029/2007JF000769 Sensors 2010,10 6999 Saleh, K ; K err, Y H ; Richaum e, P.; Escorihuela, M J.; Panciera, R.; D elw art, S.; Boulet, G ; M aisongrande,P.;W alker,J.P.;W ursteisen,P.;W igneron,J.P.Soilm oisture retrievals atL-band using a tw o-step inversion approach.Rem ote.Sens.Environ.2009,113,1304-1312 Panciera,R.; W alker,J.P.; K alm a,J.D ; K im ,E.J.; Saleh,K ; W igneron,J.P Evaluation of the SM O S L-M EB passive m icrow ave soilm oisture retrievalalgorithm Rem ote.Sens.Environ.2009, 113,435-444 K err,Y H ;W aldteufel,P.;W igneron,J.P.;M artinuzzi,J.M ;Font,J.;Berger,M Soilm oisture retrievalfrom space:The soilm oisture and O cean Salinity (SM O S)m ission.IEEE Trans.G eosci Rem ot.Sen.2001,39,1729-1735 Entekhabi, D ; N joku, E.G ; H ouser, P The H ydrosphere State (H ydros) m ission: A n Earth system pathfinderforglobalm apping ofsoilm oisture and land freeze/thaw IEEE Trans.G eosci Rem ot.Sen.2004,42,2184-2195 Jackson, T.J.; Le V ine, D M ; H su, A Y ; O ldak, A ; Starks, P.J.; Sw ift, C.T.; Isham , J.D ; H aken,M Soilm oisture m apping atregionalscales using m icrow ave radiom etry: The Southern G reatPlains hydrology experim ent.IEEE Trans.G eosci.Rem ot.Sen.1999,G E-37,2136-2151 O w e,M ;D e Jeu,R.A M ;W alker,J.P.A m ethodology for surface soilm oisture and vegetation optical depth retrieval using the m icrow ave polarization difference index IEEE Trans G eosci Rem ot.Sen.2001,39,1643-1654 10 W en,J.;Su,Z.;M a,Y D eterm ination ofland surface tem perature and soilm oisture from tropical rainfall m easuring m ission/m icrow ave im ager rem ote sensing data.J.G eophys.Res.2003,108, 4038-4050 11 Pardé, M ; W igneron, J.P.; W aldteufel, P.; K err, Y H ; Chanzy, A ; Søbjaerg, S.S.; Skou, N N -param eterretrievalfrom L-band m icrow ave observations acquired overa variety ofcrop fields IEEE Trans.G eosci.Rem ot.Sen.2004,42,1168-1178 12 W igneron,J.P.;K err,Y ;W aldteufel,P.;Saleh,K ;Escorihuela,M J.;Richaum e,P.;Ferrazzoli, P.; D e Rosnay, P.; G urney, R.; Calvet, J.C.; G rant, J.P.; G uglielm etti, M ; H ornbuckle, B.; M ätzler, C.; Pellarin, T.; Schw ank, M L-band m icrow ave em ission of the biosphere (L-M EB) m odel:D escription and calibration againstexperim entaldata sets over crop fields.Rem ote.Sens Environ.2007,107,639-655 13 W igneron, J.P.; Chanzy, A ; Calvet, J.C.; Bruguier, N A sim ple algorithm to retrieve soil m oisture and vegetation biom ass using passive m icrow ave m easurem ents over crop fields Rem ote.Sens.Environ.1995,51,331-341 14 -DFNVRQ 7- 2¶1HLOO 3( $WWHQXDWLRQ RI VRLl m icrow ave em ission by corn and soybeans at1.4 and G H z.IEEE Trans.G eosci.Rem ot.Sen.1990,28,978-980 15 W igneron, J.P.; Calvet, J.C.; D e Rosnay, P.; K err, Y ; W aldteufel, P.; Saleh, K ; Escorihuela, M J.; K ruszew ski, A Soil m oisture retrievals from biangular L-band passive m icrow ave observations.IEEE G eosci.Rem ote.Sens.Lett.2004,1,277-281 16 V an de G riend,A A ;W igneron,J.P.The b-factor as a function of frequency and canopy type at H -polarization.IEEE G eosci.Rem ote.Sens.Lett.2004,42,1-10 17 D rusch,M ; W ood,E.F.; G ao,H ; Thiele,A Soil m oisture retrieval during the Southern G reat Plains H ydrology Experim ent1999:A com parison betw een experim entrem ote sensing data and operationalproducts.W ater Resour.Res.2004,40,W 2504 Sensors 2010,10 7000 18 Cashion,J.; Lakshm i,V ; Bosch,D ; Jackson,T.J.M icrow ave rem ote sensing of soil m oisture: Evaluation ofthe TRM M m icrow ave im ager(TM I)satellite forthe Little RiverW atershed Tifton, G eorgia.J.H ydrol.2005,307,242-253 19 Bindlish, R.; Jackson, T.J.; G asiew ski, A ; Stankov, B.; K lein, M ; Cosh, M H ; M ladenova, I.; W atts,C.;V ivoni,E.;Lakshm i,V ;Bolten,J.;K eefer,T.A ircraftbased soilm oisture retrievals underm ixed vegetation and topographic conditions.Rem ote.Sens.Environ.2008,112,375-390 20 Jackson, T.J.; Schm ugge, T.J V egetation effects on the m icrow ave em ission of soils Rem ote Sens.Environ.1991,36,203-212 21 Schm ugge, T.J.; Jackson, T.J A dielectric m odel of the vegetation effects on the m icrow ave em ission from soils.IEEE Trans.G eosci.Rem ot.Sen.1992,30,757-760 22 Bindlish,R.;Jackson,T.J.;W ood,E.;G ao,H ;Starks,P.;Bosch,D ;Lakshm i,V Soilm oisture estim ates from TRM M M icrow ave Im agerobservations overthe Southern U nited States.Rem ote Sens.Environ.2003,85,507-515 23 Saleh,K ;W igneron,J.P.;D e Rosnay,P.;Calvet;J.C.;Escorihuela,M J.;K err,Y ;W aldteufel,P Im pact of rain interception by vegetation and m ulch on the L-band em ission of natural grass Rem ote.Sens.Environ.2006,101,127-139 24 M o, T.; Choudhury, B.J.; Schm ugge, T.J.; W ang, J.R.; Jackson, T.J A m odel for m icrow ave em ission from vegetation-covered fields.J.G eophys.Res.1982,87,11229-11237 25 W ang,J.R.;Choudhury,B.J.Rem ote sensing of soilm oisture contentover bare field at1.4 G H z frequency.J.G eophys.Res.1981,86,5277-5282 26 :DQJ -5 2¶1HLOO 3( -DFNVRQ 7- (QJPDQ (7 0XOWLIUHTXHQF\ PHDVXUHPHQWV RI WKH effects of soil m oisture, soil texture, and surface roughness IEEE Trans G eosci Rem ot Sen 1983,G E-21,44-51 27 Lang, R.H ; Sidhu, J.S Electrom agnetic backscattering from a layer of vegetation: A discrete approach.IEEE Trans.G eosci.Rem ot.Sen.1983,21,62-71 28 Chauhan, N S Soil m oisture estim ation under a vegetation cover: Com bined active passive m icrow ave rem ote sensing approach.Int.J.Rem ote.Sens.1997,18,1079-1097 29 N joku, E.G A M SR land surface param eters: Surface soil m oisture, land surface tem perature, vegetation w ater content; A lgorithm Theoretical Basis D ocum en, Jet Propulsion Laboratory, California Institute ofTechnology,Pasadena,CA ,U SA ,1999 30 K err, Y H ; W aldteufel, P.; Richaum e, P.; D avenport, I.; Ferrazzoli, P.; W igneron, J.P SM O S level processor soil m oisture A lgorithm Theoretical Basis D ocum ent (A TBD ) SM -ESL (CBSA ),CESBIO ,Toulouse,SO -TN -ESL-SM -G S-0001,V 5.a,15/03/2006,2006 31 G ish, T.J.; W althall, C.L.; D aughtry, C.S.T.; D ulaney, W P.; M ccarty, G W W atershed-Scale Sensing of Subsurface Flow Pathw ays at O pe3 Site In Proceedings of the First Interagency Conference on Research in the W atersheds, Benson, A Z, U SA O ctober 27-30, 2003; pp.192-197 32 -RVHSK$7 9DQGHU9HOGH5 2¶1HLOO3( /DQJ5+ *LVK76RLOPRLVWXUHUHWULHYDOGXULQJ a corn grow th cycle using L-band (1.6 G H z)radarobservations.IEEE Trans.G eosci.Rem ot.Sen 2008,46,2365-2374 33 M iller,J.D ;G askin,G D Theta probe M L2x²Principle ofoperation and applications,2nd ed.; M acaulay Land U se Research Institute (M LU RI).Crairiebuckler,A berdeen,U K ,1996 Sensors 2010,10 7001 34 Cosh.,M H ;Jackson,T.J.;Bindlish,R.;Fam iglietti,J.S.;Ryu,D Calibration of an im pedance probe forestim ation ofsurface soilw atercontentoverlarge regions.J.H ydrol.2005,311,49-58 35 D obson,M C.;U laby,F.T.;H allikainen,M T.;El-Rayes,M A M icrow ave dielectic behavior of w et soil²Part II: D ielectric m ixing m odels IEEE Trans G eosci Rem ot Sen 1985, G E-23, 35-46 36 Choudhury,B.J.;Schm ugge,T.J.;N ew ton,R.W ;Chang,A T.C.Effectof surface roughness on m icrow ave em ission ofsoils.J.G eophys.Res.1979,84,5699-5706 37 V an de G riend,A A ; W igneron,J.P.O n the m easurem ent of m icrow ave vegetation properties: Som e guidelines fora protocol.IEEE Trans.G eosci.Rem ot.Sen.2004,42,2277-2289 38 W igneron, J.; Laguerre, L.; K err, Y A sim ple param eterization of the L-band m icrow ave em ission from rough agriculturalsoils.IEEE Trans.G eosci.Rem ot.Sen.2001,39,1697-1707 39 Escorihuela,M J.;K err,Y H ;D e Rosnay,P.;W igneron,J.P.;Calvet,J.C.;Lem tre,F.A sim ple m odelofthe bare soilm icrow ave em ission atL-band.IEEE Trans.G eosci.Rem ot.Sen.2007,45, 1978-1987 40 Choudhury, B.J Reflectivities of selected land-surface types at 19 and 37 G H z from SSM /I observations.Rem ote.Sens.Environ.1993,46,1-17 41 Le V ine,D M ;K aram ,M A D ependence ofattenuation in a vegetation canopy on frequency and plantw atercontent.IEEE Trans.G eosci.Rem ot.Sen.1996.34,1090-1096 © 2010 by the authors; licensee M D PI, Basel, Sw itzerland This article is an O pen A ccess article distributed under the term s and conditions of the Creative Com m ons A ttribution license (http://creativecom m ons.org/licenses/by/3.0/) Copyright of Sensors (14248220) is the property of MDPI Publishing and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission However, users may print, download, or email articles for individual use ... ation ofthe Ȗ from m icrow ave data is preferred because the ancillary data needed ata globalscale forsoilm oisture retrievalm ay notbe available.H ow ever,its dependence on the instrum entparam... utilized data set has been collected atone of the fields of the Beltsville A griculturalResearch Center (BA RC) by an autom ated tow erm ounted L- band (1.4 G H z)radiom eter(called LRA D )starting... sub-surface soil and w ater flow characteristics and covers an area of 25 near Beltsville,M aryland (Figure 1).Each ofthe fourw atersheds is form ed from sandy fluvialdeposits and has a varying slope