Scientia Horticulturae 137 (2012) 4958

Contents lists available at SciVerse ScienceDirect

Scientia Horticulturae

journa l h o me page: www.elsev ier .com

Determ e hinfrare

A. LpezUniversidad de s/n, 0

a r t i c l

Article history:Received 12 MReceived in reAccepted 18 Ja

Keywords:Infrared thermCrop emissivitCanopy tempeHorticultural c

h the l cropused

the late ed devg avem ananum06 fo

L. Considerable differences have been observed between the emissivity values on the opposite sides ofthe leaves in some horticultural crops, such as green bean and particularly red bean, with a differenceof 0.029 in the average emissivity value. Emissivity values of 0.98 are recommended as a reference formeasuring the temperature of horticultural crops other than those studied here whenever there is noother possibility to determine the emissivity.

2012 Elsevier B.V. All rights reserved.

1. Introdu

Measuriraphy is beof experimculture. Wthermometcrop tempecontinuousYeater, 200ture was esand Rosenbwas initiallthermocoupNilsson, 199since theseand therefoable by direserious proa meaningf

CorresponE-mail add

0304-4238/$ doi:10.1016/j.ction

ng the temperature of objects by infrared thermog-coming more and more frequent in a wide variety

ental elds, amongst which we should mention agri-hile air temperature is quite easy to measure withers, thermocouples or thermistors, the measurement ofrature is usually more difcult to achieve, mainly when

non-contact measurements are necessary (Mahan and8). The importance of plant temperature in agricul-tablished towards the late 1970s and early 1980s (Bladurg, 1976; Thofelt, 1977; Idso, 1982; Jackson, 1982). Ity studied by measuring leaf temperature with micro-les (Hurd and Bailey, 1983; Caouette et al., 1990;1), but this technique can be inefcient (Tanner, 1963)

devices easily become detached (Meyer et al., 1994)re record incorrect data. The measurement of this vari-ct contact sensors presents several problems, the mostbably being the need for spatial integration to obtainul average (Fuchs and Tanner, 1966). Most of these

ding author. Tel.: +34 950015546; fax: +34 950015491.ress: dvalera@ual.es (D.L. Valera).

problems can be overcome by measuring the thermal radiationemitted by the canopy as a whole (Berliner et al., 1984). Con-sequently, we have considered non-contact measurement of leaftemperature a more suitable technique for monitoring the temper-ature of vegetable crops.

Radiometric surface thermometers or infrared thermometers(IRTs) can be used to measure crop temperature. Infrared ther-mometers with a band pass lter from 8 to 13 m allowedmeasurement of the real temperature of plant surfaces with errorsin the range of 0.10.3 C (Fuchs and Tanner, 1966). The advantagesof this method include the fact that there is no need for physi-cal contact with the plant, simple automation of data collectionand non-point measurements that accommodate inherent spatialvariability (Mahan and Yeater, 2008).

The above technique has been used to study crop temperaturein many recent studies, for example to analyse the relationshipbetween leaf temperature and water use and growth of plants(Tanner, 1963; Jackson et al., 1981; Hateld et al., 1983; Choudhuryet al., 1986; Hateld, 1990; Wanjura and Mahan, 1994; Pinter et al.,2003; Peters and Evett, 2004). The canopy temperature (as derivedfrom thermal radiation measurements) and notably its relation-ship with selected reference variables have been used as a basis fordening stress indices (Aston and Van Bavel, 1972; Idso et al., 1981;Jackson et al., 1981). Guimaraes et al. (2010) evaluated the use of

see front matter 2012 Elsevier B.V. All rights reserved.scienta.2012.01.022ining the emissivity of the leaves of nind thermography

, F.D. Molina-Aiz, D.L. Valera , A. Pena Almera, Campus de Excelencia Internacional Agroalimentario ceiA3. Ctra. Sacramento

e i n f o

ay 2011vised form 16 January 2012nuary 2012

ographyyraturerops

a b s t r a c t

The present study was carried out witof the most characteristic horticulturaregion. A thermographic camera was by evaluating radiation emission fromwith a contact probe in order to calcuupper side of leaves are below standarside of leaves we obtained the followinlentum Mill., 0.978 0.008 for CapsicuCucurbita pepo L., 0.973 0.007 for Solfor Citrullus lanatus Thunb., 0.983 0.0/ locate /sc ihor t i

orticultural crops by means of

4120 Almera, Spain

aim of analysing the variability of the emissivity values of nines of the greenhouse productive system in the Mediterranean

for both qualitative and quantitative emissivity measurementleaves. The real temperature of the leaves was also measuredmissivity. The differences in emissivity between crops for theiation values, the average values are all close to 0.98. For upperrage values of emissivity: 0.980 0.010 for Lycopersicum escu-nuum L., 0.983 0.008 for Cucumis sativus L., 0.985 0.007 for

melongena L., 0.978 0.006 for Cucumis melo L., 0.981 0.009r Phaseolus vulgaris L. and 0.983 0.005 for Phaseolus coccineus

50 A. Lpez et al. / Scientia Horticulturae 137 (2012) 4958

Nomenclature

Rh energy ux emitted by leaves, W m2

RTT TaTh0.98

TreTsTsub

Tw

refh

infrared theintra speci

Model sito investigaestimation optimal meenergy balabiometeororespiration ature (Leuzithis respectbe dissipateperature toclosure cauinfrared theshowed a coture using asynthesise tance of a lreference le

The signplant waterand irrigatiohas been usand TannerGreenhouseleaf temperture and traclimate (Hamethods haductance frLeinonen et

Radiomeera measurtemperaturband pass oFor accurattion, the ema correctionroundings 1996). The ean object at

emitter (Husehke, 1959). The values for emissivity may range fromzero to unity. If an incorrect value is assumed, error must result(Hipps, 1989). Several eld methods were developed to determinecanopy emissivity in the infrared region (Fuchs and Tanner, 1966;

er anand hf lan

Hum stud(198ata L. Thean in

Rahr Brance tat mes, or hels.en thf ho

out hara

regirface

estim

teria

eore

radiourcer obc con

T4 +

RT is, is is Stecorrehere

temraturradiance entering a thermographic camera, W m2

object temperature, Kair temperature, Kleaf temperature measured with a thermographiccamera with reference emissivity, Kreected temperature, Kleaf temperature measured with a contact probe, Kwater temperature measured with a submergedprobe in the bath, Kwater temperature measured with a thermographiccamera, Kwavelength, mStefanBoltzmanns constant,5.67051 108 W m2 K4emissivityreference emissivity equal to 0.98calculated emissivity of the leavesspectral transmittance of atmosphere

rmometry in the characterisation of inter specic andc upland rice lines for drought tolerance.mulations and experimental measurements were usedte the applicability of infrared thermography for theof stomatal conductance and drought stress under sub-teorological conditions (Maes et al., 2011). The leafnce and resulting leaf temperature are central themes oflogy; transpiration, sensible heat ux, photosynthesis,and other metabolic activities are driven by leaf temper-nger and Krner, 2007). The stomata play a major role in, as on closing they limit the amount of energy that cand by transpiration and consequently cause the leaf tem-

increase (Raschke, 1960). Drought-induced stomatalses a rise in canopy temperature that can be detected byrmometers (Ehrler et al., 1978). Hashimoto et al. (1981)rrelation between leaf temperature and stomatal aper-

thermal camera, and Guilioni et al. (2008) clarify andthe appropriate equations linking the stomatal resis-eaf to its own temperature and to the temperatures ofaves (dry and wet).icance of leaf temperature and stomatal aperture for

relations was acknowledged in the early 20th century,

Buettnsound sivity o1991;

FewHipps trident(1966)ibrate plants.0.98 foemittaof a grefor treherbs dry soi

Givsivity ocarriedmost craneanthe su(e.g. to

2. Ma

2.1. Th

Thethree s(ii) othspheri

RT =

whereW m2

ter), (1 ) atmospgroundtempen scheduling by means of canopy temperature surveysed in agriculture and horticulture since the 1960s (Fuchs, 1966; Jackson et al., 1977; Fuchs, 1990; Jones, 2004).

researchers have long been interested in measuringatures for the purposes of estimating stomatal aper-nspiration at canopy level, and controlling greenhouseshimoto et al., 1981; Bakker, 1984; Ehret, 2001). Someve been developed to estimate accurately stomatal con-om leaf temperatures (Jones, 1999; Jones et al., 2002;

al., 2006).tric surface temperatures obtained from thermal cam-ements are a function of both the physical surfacee and the effective emissivity of the surface within thef the radiometric measurement (Humes et al., 1994).

e measurement of crop temperature by infrared radia-issivity must either be known or determined, making

accounting for the reected radiation from the sur-(Fuchs and Tanner, 1966; Hipps, 1989; Sugita et al.,missivity, , describes the ratio of radiation emitted by

a certain temperature, to the value emitted by a perfect

Eq. (1) perature mbackgroundsurements Bartholic (1et al. (1990

2.2. Experim

The expjust separat

Thermographic camand the watoring of air

2.2.1. IR theThe ther

pact infrareDanderyd, d Kern, 1965). Most of these methods are physicallyave been widely used for the determination of the emis-d surfaces (Blad and Rosenburg, 1976; Labed and Stoll,es et al., 1994).ies have been carried out to determine crop emissivity.

9) obtained a value of emissivity of 0.97 for Artemisia. following the method described by Fuchs and Tanner

same method was used by Berliner et al. (1984), to cal-frared thermometer used to estimate water stress inkonen and Jokela (2003) determined an emissivity ofssica rapa L. and Sonchus arvensis L. with a referenceechnique. Rubio et al. (1997) calculated the emissivityany varieties, nding average values of = 0.983 0.004

= 0.984 0.007 for shrubs, = 0.984 0.009 for wetrbs on wet soils and = 0.962 0.013 for dry herbs or

e shortage of previous studies determining the emis-rticultural crops such as tomato, the present study waswith the aim of making known emissivity values of thecteristic crops of the productive system in the Mediter-on. These values of emissivity are required when using

temperature of crops in energy balance applicationsate stomatal conductance, or to assess drought stress).

ls and methods

tical considerations

ance entering a thermographic camera originates fromes (Lamprecht et al., 2002): (i) the observed object itself;jects reected on the targets surface, and; (iii) an atmo-tribution.

(1 )T4refl + (1 )T4a (1)

the energy ux emitted at a wavelength of 7.313 m inthe emissivity of the target (equal to 1 for a perfect emit-fanBoltzmanns constant (5.67051 108 W m2 K4),sponds to the reectivity, (1 ) is the emittance of the, T is the temperature of the target, Tre is the back-perature that the target is reecting and Ta is the aire, all in K.addresses the two sources of error in radiative tem-easurements, namely the estimates of emissivity and

temperature. These errors in plant temperature mea-have been discussed in more detail by Sutherland and979), Amiro et al. (1983), Hipps (1989) and Svendsen).

ental arrangement

erimental conguration is shown in Fig. 1. Fresh leaves,ed from a plant, were placed oating in a water bath.metry was performed in two ways: (i) using a thermo-era to determine the surface temperatures of leaves

ter; and (ii) using data loggers for a continuous moni-, water and leaf temperatures.

rmographymographic images (Fig. 2) were recorded with a com-d camera ThermoVisionTM A40-M (FLIR Systems AB,

Sweden), with a spectral infrared range of wavelength

A. Lpez et al. / Scientia Horticulturae 137 (2012) 4958 51

Fig. 1. Experimental arrangement for measuring emissivity of one side of a plantleaf by imaging infrared thermography while the other side is oating in a waterbath.

from 7.3 to 13 m, a temperature range of 40 to +120 C and anaccuracy of 2%. The detector was a Focal Plane Array, uncooledmicrobolometer of 320 240 pixels and the eld of view was24 18 with a minimal focus distance of 0.3 m. The spatial reso-lution was 0.08 C at 30 C.

The caThermaCAMDanderyd, such as poideterminatrange of thimages can

2.2.2. ThermThe anal

of the tempthe object (the reectetivity humiHOBO ProUSA) with awater was (Onset Com+20 C, besi

Fig. 2. Analysiselected areasto measure the

For a correct calculation of the emissivity the real tempera-ture of the leaf must be known at each moment. As a result, aswell as the surface temperature of the water measured by thethermographic camera and the water temperature measured withthe probe ssured withBarcelona (ment rangestudy whetthermograp

The star45 C (the tevals of 2.5sequences two sequenleaves. For different mof the plantmaterial wtotal of 180underside o

2.2.3. Leaf e emmittahkonientratured aater965;986;as m

surfaater

n theratur

follovege

the raturecord of

(Fig. (1 HaCAMarrietly, tequemera is supported by the software packageTM Researcher Pro 2.8 SR-3 (FLIR Systems AB,

Sweden), which offers numerous analysis functionsnt temperatures, proles, histograms, isotherms or theion of the maximum temperature in the image. Thee actual temperature and the false-colours of the IR

be chosen as desire.

ometryysis requires some parameters for a correct adjustmenterature values. It is necessary to know the distance of0.5 m), the temperature and humidity of the air, andd temperature. Continuous air temperature and rela-dity of the air were measured with two dataloggers

Temp-HR U23-001 (Onset Computer Corp., Pocasset,ccuracy of 0.18 C and 2.5%. The temperature of themeasured with a submerged probe HOBO TMC6-HCputer Corp., Pocasset, USA) with accuracy of 0.5 C atdes the reading of the Thermomix bath.

Theence eand Raat ambtempewas usity of wKern, 1et al., 1level wwater face (woat otempe

Theof the ing ontempeing to rleaf anstoredsecondThermwere c

Firseach ss of a thermographic image for a water temperature of 37.5 C with 7: 5 distributed in 2 aubergine leaves, 1 to dene the water bath and 1

temperature reected by the aluminium sheet.

calculate themittance othe relative (FLIR, 200

Secondlreected byside over thand acts asanalysis socarrying oureected te

Once wehumidity a(0.5 m), a swater tempof the leaveubmerged in the bath, the leaf temperature was mea- a contact probe SR-TFH-DISC, Desin Instrument, S.A.,Spain), with accuracy of 0.4 C at +20 C and a measure-

of 0150 C (Fig. 3b). These measurements were used toher the surface temperature of the water obtained fromhic images and the leaf temperature were the same.ting temperature of the bath was 25 C, increasing tomperature used by Rahkonen and Jokela, 2003) at inter-

C. For each xed temperature, we have registered 4of images of 3 min duration with a frequency of 1 Hz,ces for the upper side and two for the underside of theeach of these sequences and for each temperature, twoature leaves are collected fresh from the mid-lower part

analysed. In each sequence, 5 surfaces of the vegetableere dened to measure the temperature. This means a0 emissivity values calculated for the upper side andf each horticultural crop.

missivity measurementissivity of the leaves was measured with the refer-nce technique described by Fuchs and Tanner (1966)en and Jokela (2003). Measurement was performed

room temperatures of 22 2 C and with water bathes ranging from 25 to 45 C. A well-stirred water baths a reference material (Berliner et al., 1984). Emissiv-

( = 812 m) was assumed to be 0.98 (Buettner and Robinson and Davies, 1972; Pinkley et al., 1977; Zhang

Salisbury and Milton, 1988). The background radiationeasured by using an aluminium foil inserted near thece as a reector. The radiation levels of a reference sur-

as reference material) and a leaf carefully lowered to water were measured. This method assumed that thee of the water and the leaf is the same.wing procedure was used to determine the emissivity

table material: freshly picked leaves were placed oat-water in the bath which had been heated to the givene. A certain time (in about 1 min) was left before start-d images in order to ensure that the temperature of thethe water was the same. Thermographic images were

2) for three minutes at a frequency of one image perz). From the sequence of images obtained and using theTM Researcher Pro 2.8 SR-3 software several analyses

d out.he average values of air temperature and humidity innce of images are introduced so that the software cane fraction of radiation emitted by the atmosphere. Thef the atmosphere (1 ), which is heavily dependent on

humidity of the air. In this way the software estimates6).y, it is necessary to calculate the fraction of radiation

the leaves. A sheet of aluminium foil was placed on onee water bath. Aluminium foil has very low emissivity

a reector. By setting the emissivity to 1 in the imageftware for the area of the aluminium foil (Fig. 1) andt a rst analysis of the images, the average value of themperature is obtained (Tre).

know the reected temperature, air temperature andnd the distance from the camera to the water bathecond image analysis is carried out to determine theerature (Tw) with emissivity 0.98, and the temperatures (Th0.98) also with a reference emissivity of 0.98.

52 A. Lpez et al. / Scientia Horticulturae 137 (2012) 4958

Fig. 3. Area selected for analysis of temporal variation and the transversal line for the analysis of spatial variation (a). Thermographic image of the contact temperaturesensor on a tomato leaf (b).

From thfollowing e

Rh = ref T

where Rh is7.313 m for a referen

The totacan be expr(h), the unture of the lat the same

Rh = hT4wFrom Eq

real emissiv

h = ref T

For the(42 cm 25the water omove outsiwell-stirred(Braun Biotworking ran

erma

eral aCAMprocata, nt anissivitatgSA). en thnces.

Tempporper

xim. Spatl imaf. Le graross

Emis each

Table 1Temperature owith the subm

Tw (C) Ts (C) (Tw Ts)/Tw

Tw (C) Tsub (C) (Tw Tsub)/Te analysis carried out on the surface of the leaves, thexpression is obtained:

4h0.98 + (1 ref )T4refl + (1 )T4a (2)

the energy ux emitted by the leaves at a wavelength ofin W m2, and Th0.98 is the temperature of the target,ce emissivity of ref = 0.98.

l radiation reaching the camera focussing on the leavesessed as a function of the real emissivity of the leavesknown that is the object of study, and the real tempera-eaves, assuming that the water bath and the leaves are

temperature, which would be Tw:

+ (1 h)T4refl + (1 )T4a (3)

s. (2) and (3) we obtain the expression which allows theity of the leaves, h, to be calculated:

4h0.98 T4reflT4w T4refl

(4)

water bath a rectangular plastic container cm 16 cm) was used. The leaves stay oating inn a metallic grill, which ensures that the leaves do notde the cameras focal range. The temperature of the

water was controlled with a Thermomix BM agitatorech International, Melsungen, Germany), which has a

2.3. Th

SevThermimage ment ddiffere

Emusing SMD, UbetweDiffereof 95%

2.3.1. Tem

age temat a masivity)thermacrop leing lina line c

2.3.2. Forge of 22100 C and an accuracy of 0.03 C. of four diffe

f the water bath calculated from the thermographic images (Tw); temperature of the leaveerged sensor (Tsub).

Thermomix BM temperature setpoint (C)

40.0 40.0 45.0

37.90 37.71 42.06 37.79 37.50 41.88

(%) 0.29 0.56 0.43

Thermomix BM temperature setpoint (C)

25.0 27.5 30.0 32.5 35.0

25.18 27.59 29.70 31.93 34.1625.16 27.52 29.94 32.74 35.17

w (%) 0.08 0.25 0.81 2.54 2.96l image analysis

different thermal analyses were performed using theTM Researcher Pro 2.8 SR-3 digital infrared thermal

essing software. Due to the large amount of measure-not all the material was included in each analysis. Thealyses and the materials used in each are listed below.ty data were subjected to Analysis of Variance (ANOVA)raphics Plus 4.1 Software (Manugistics, Inc., Rockville,One-way ANOVA and possible signicant differencese emissivity values were evaluated by Least Signicant

(LSD) multiple comparison tests with a condence level

oral and spatial variation in temperatureal variation in temperature was analysed by the aver-ature of a selected area (Fig. 3a) of a leaf versus timeum sample rate of 1 Hz over 3 min (180 values of emis-ial variation in temperature was analysed by capturingages at ve moments during the experiments for everyinear temperature distributions were studied by creat-phs (Fig. 3a) describing temperatures (80 values) alonging a leaf.

sivity variation amongst leaves crop, the emissivity values of the upper and lower sides

rent groups of leaves were determined independently

s measured with the contact sensor (Ts); water temperature measured

45.0 45.0 45.0

41.93 42.08 42.0841.84 42.03 42.050.21 0.12 0.07

37.5 40.0 42.5 45.0

36.35 38.57 40.77 42.79 37.44 40.13 42.46 44.89 3.00 4.04 4.15 4.91

A. Lpez et al. / Scientia Horticulturae 137 (2012) 4958 53

Fig. 4. Emissivfor the unders

for bath temused ve lea(900 valuesanalysis to of leaves.

2.4. Vegeta

The hor(LycopersicuL.), cucumbaubergine (melon (Citrand red bea

3. Results and discussion

Firstly we have studied the difference between the temperatureof the leaves oating in the water bath, obtained by the thermo-graphic camera, and the temperature of the water as detected bythe sensor and by the thermographic camera. This was followed bya study of the temporal and spatial variation of the leaf temperatureduring the assays. Finally the emissivity for the nine horticul-tural species was calculated. Differences were detected betweenspecies and between the upper and lower sides of the leaf of givenspecies, and the results obtained were compared with the valuesof emissivity found in the literature for each type of vegetablematter.

3.1. Difference between leaf and water temperature

The values of leaf temperature, measured with the contactsensor, and water temperature, obtained by the thermographiccamera, were very similar (Table 1). The maximum differenceobserved was 0.21 C. In all cases the differences detected were lessthan the accuracy margin of the thermographic camera (2%). If theleaves are maintained oating on the water bath at a constant tem-

re for a prudent length of time (35 min) before commencingay, ired bissivce m

sligcultrface

the a

ther is thof soe senisc cleaveity values calculated for the upper side of cucumber leaves (a) andide of courgette leaves (b).

peratures from 35 C to 45 C (for each group have beenves, one for each water bath temperature). These values

for each group of leaves) then underwent statisticaldetermine the variation in emissivity amongst groups

ble materials

peratuthe assmeasuthe emreferen

Thethe difleaf sunot allleaves.

Anosensorleaves tion, thmetal dto the ticultural crops studied in this work were tomatom esculentum Mill.), green pepper (Capsicum annuumer (Cucumis sativus L.), courgette (Cucurbita pepo L.),Solanum melongena L.), melon (Cucumis melo L.), water-ullus lanatus Thunb.), green bean (Phaseolus vulgaris L.)n (Phaseolus coccineus L.).

sible. It waleaf was indid not getimpossible other occaswere found

Fig. 5. Images of the upper side of aubergine leaves oating on the batht can be assumed that the leaf and water temperaturey the thermographic camera is the same, and thereforeity of the leaves can be determined using the water asaterial.

ht differences in temperature observed may be due toy in placing the contact sensor correctly on the irregular. It is precisely this difculty that is the main reason whyssays were carried out with the contact sensor on the

drawback found when using the contact temperaturee ne layer of hairs on the top and/or underside of theme of the crops studied, for instance aubergine. In addi-sor is not easy to handle. It consists of a 15 mm diameteronnected to a PC by a rather inexible cable. It was xeds using adhesive tape to ensure the best contact pos-s then placed on the water bath, taking care that the

full contact with the water, but also that the sensor wet. The rigidity of the cable made this task almostand many assays were invalidated because of this. Onions, once the assays had been completed, small ssures

to have been made in the leaves by the sensor disc,

of water at 25 C (a) and 37.5 C (b).

54 A. Lpez et al. / Scientia Horticulturae 137 (2012) 4958

Table 2Emissivity values (average value standard deviation) of the upperside of the leaves of nine horticultural crops obtained by the method of the reference material for 9different water bath temperatures.

Upper side Thermomix BM temperature setpoint (C)

35.0 37.5 40.0 42.5 45.0

Tomato 0.978 0.010 0.980 0.009 0.979 0.009 0.982 0.007 0.982 0.011Pepper 0.978 0.008 0.977 0.008 0.977 0.007 0.978 0.008 0.977 0.007Cucumber 0.982 0.010 0.982 0.006 0.982 0.007 0.986 0.007 0.984 0.010Courgette 0.986 0.007 0.985 0.007 0.985 0.007 0.986 0.008 0.983 0.007Aubergine 0.971 0.006 0.971 0.007 0.974 0.007 0.973 0.008 0.975 0.007Melon 0.978 0.006 0.978 0.006 0.980 0.006 0.979 0.005 0.977 0.007Watermelon 0.982 0.009 0.981 0.009 0.980 0.008 0.978 0.010 0.982 0.009Green bean 0.982 0.007 0.984 0.006 0.983 0.024 0.981 0.006 0.983 0.006Red bean 0.983 0.005 0.983 0.005 0.982 0.005 0.983 0.005 0.985 0.005

meaning that the sensor had got wet, and these assays were alsofruitless.

To determine the emissivity of the leaves we have used thesurface temperature of the water instead of the underwatertemperaturter temperasensible heaverage temera and by with peppefrom 25 tofrom 0.08% peratures mthe set-pointor. We canwater bath water temppoint tempthe leaves.

3.2. Tempo

The higobserved atvery close ttion, the lea

The spatthe greateswas observand the low

Althougporal variatemperaturthe accurac

3.3. Emissivity of the crops

Tables 2 and 3 show the emissivity values obtained at differentwater bath temperatures.

konet 45e prer to ae h

mpe greaariatose tradiaiatio

phic tes tat a s alsotion o am

watvity whicor thtte le

highted eAlso r ima

greallowve onraturratur

Table 3Emissivity val erent for 9 different

Underside

Tomato Pepper Cucumber Courgette Aubergine MelonWatermelonGreen beanRed bean e. The surface temperature is less than the underwa-ture, as it could be affected by evaporative cooling andat exchange with the environment. Table 1 shows theperature values recorded by the thermographic cam-the submerged temperature sensor during the assaysr leaves. As the temperature of the water bath increases

45 C, so does the difference between both increasesto 4.91%. As we can observe in Table 1 the water tem-easured with the submerged sensor are very close tot temperatures showed by the Thermomix BM agita-

also observe that these set-point temperatures of theare greater than the leaf temperatures and the surfaceeratures. So is very important to emphasise that the set-erature is not appropriate as reference temperature for

ral and spatial temperature variation in leaves

hest temporal variation in leaf temperature was the lowest water bath temperature (25 C), which iso ambient temperature. In the cases of greatest varia-f temperature was 25.22 0.12 C.ial temperature variation was acceptable for all crops,t spatial variation in leaf temperature (40.72 0.15 C)ed for an image taken with the water bath at 42.5 C,est (34.16 0.08 C) at 35 C.h the spatial variation was slightly greater than the tem-tion, in both cases the dispersion or variation in leafe obtained by this method is acceptable, being less thany margin of the thermographic camera.

Rahwater aL. In thin ordetions. Wbath tewith aature vwas cltity of the radmograillustrataken authorcalibraclose t

ForemissiFig. 4, lated fcourge

Forcalculavalue. and fo

Thetures awe hatempetempe

ues (average value standard deviation) of the underside of the leaves of nine diff

water bath temperatures.

Thermomix BM temperature setpoint (C)

35.0 37.5 40.0

0.982 0.008 0.977 0.008 0.980 00.983 0.007 0.985 0.006 0.992 00.985 0.010 0.986 0.009 0.985 00.984 0.008 0.986 0.007 0.981 00.966 0.008 0.966 0.009 0.969 00.984 0.008 0.981 0.005 0.981 0

0.986 0.011 0.986 0.009 0.985 00.976 0.017 0.969 0.009 0.965 00.951 0.012 0.954 0.008 0.957 0n and Jokela (2003) set the temperature of the bath ofC to determine the emissivity of B. rapa L. and S. arvensissent work we have used a wider range of temperaturesnalyse the inuence of temperature on emissivity varia-ave observed that the emissivity values calculated usingratures close to ambient temperature are not accuratet dispersion (Fig. 4), and greater temporal leaf temper-ion is also observed. When the temperature of the batho ambient temperature (25 C and 27.5 C), the quan-tion emitted by the vegetable material was similar ton emitted by the water. In these conditions, the ther-camera does not obtain a clear image of leaves. Fig. 5he difference in clarity between a thermographic imagewater temperature of 25 C and one at 37.5 C. Other

observed dispersion in temperature values during theof infrared thermometers with water bath temperaturesbient values (Churchill et al., 1982; Berliner et al., 1984).er temperature close to air temperature we obtainedclose to 1 for all crops analysed, as can be observed inh shows the dispersion of the emissivity values calcu-e upper side of cucumber leaves of and the underside ofaves.er bath temperatures, the dispersion of the values ofmissivity is very low, as all are very near the averagethese values were very similar between temperaturesges with the same water temperature (Fig. 4).ter radiation emitted by the material at these tempera-s a correct calculation of the emissivity. For this reason,ly considered the emissivity values obtained for bathes from 35 C to 45 C. It is recommendable to use bathes that are at least 15 C above the ambient temperature.

horticultural crops obtained by the method of the reference material42.5 45.0

.007 0.985 0.011 0.979 0.010

.007 0.994 0.007 0.983 0.006

.009 0.988 0.008 0.986 0.008

.008 0.985 0.006 0.982 0.008

.009 0.968 0.008 0.970 0.008

.004 0.979 0.005 0.978 0.005

.008 0.983 0.009 0.984 0.008

.012 0.967 0.016 0.961 0.015

.013 0.952 0.008 0.958 0.012

A. Lpez et al. / Scientia Horticulturae 137 (2012) 4958 55

Fig. 6. eaves

The avercorrespondbetween 7.3

The emiunder study(Table 5). Bity of = 0.obtained emmethod estJokela (200sis L. ( = 0.9well-stirred(1997) calcaverage valshrubs, = 0 = 0.962 0

3.4. Emissiv

Althoughfound betwon the who0.001. This vof three crodifference rysis indicat

ty valalysedraphi

or ber

ette gine

Cucumis melo L. 0.978 0.006 0.980 0.006melon Citrullus lanatus Thunb. 0.981 0.009 0.985 0.009

Bean Phaseolus vulgaris L. 0.983 0.006 0.968 0.015an Phaseolus coccineus L. 0.983 0.005 0.954 0.011

ments provides sufcient accuracy in obtaining the emissiv-es.

fference between the upper side and the underside of the

ato was the only vegetable for which no statistical differ-ere found between the emissivity on opposite sides of the

g. 6a). The results of other vegetables in ascending order ofStatistical analysis of the emissivity values of the upper side and underside of the l

age values of emissivity obtained in the present study to the spectral range of the thermographic camera,

and 13 m.ssivity values obtained (Table 4) for most of the crops

are very similar to those considered by other authorsrewster (1992) and Meyer et al. (1994) use emissiv-98 as generic for different vegetables. Hipps (1989)issivity of = 0.97 for A. tridentata L. following the

ablished by Fuchs and Tanner (1966). Rahkonen and3) determined the emissivity of B. rapa L. and S. arven-8) with a reference emittance technique also using a

water bath at a temperature of about 45 C. Rubio et al.ulated the emissivity of a great many varieties, ndingues of = 0.983 0.004 for trees, = 0.984 0.007 for.984 0.009 for wet herbs or herbs on wet soils and.013 for dry herbs or dry soils.

ity variation amongst leaves

in some crops statistically signicant differences wereeen the emissivity of different groups of leaves (Table 6),le the maximum differences recorded were less than

Table 4Emissivicrops anthermog

TomatPeppeCucumCourgAuberMelonWaterGreenRed Be

experiity valu

3.5. Dileaves

Tomences wleaf (Fialue was only surpassed for the underside of the leavesps (tomato, courgette and red bean), and the maximumecorded was 0.0018 for courgette. This statistical anal-es that the number of leaves and replications in the

differencescourgette, cwere signithe emissiv

Fig. 7. Leaves of red bean oating in the water bath: upper sof tomato (a), cucumber (b), watermelon (c) and red bean (d).

ues (average value standard deviation) for the leaves of the nine calculated using the temperature of the water measured with the

c camera (from 35 to 45 C).

Crops Upper side Underside

Lycopersicum esculentum Mill. 0.980 0.010 0.981 0.009Capsicum annuum L. 0.978 0.008 0.987 0.008Cucumis sativus L. 0.983 0.008 0.986 0.009Cucurbita pepo L. 0.985 0.007 0.984 0.008Solanum melongena L. 0.973 0.007 0.968 0.008 in emissivity between the opposite sides of the leaf wereucumber (Fig. 6b) and melon, for which the differencecant (0.003) but less than the standard deviation ofity values calculated. A second group consisted of water-

ide (a) and underside (b).

56 A. Lpez et al. / Scientia Horticulturae 137 (2012) 4958

Table 5Emissivity values obtained by several authors for different plants.

Crops Emissivity Source

Snap bean Phaseolus vulgaris L. 0.96 Fuchs and Tanner (1966)Tobacco 0.97 Fuchs and Tanner (1966)Artemisia 0.97 Hipps (1989)Alfalfa 0.970.98 Fuchs and Tanner (1966)Sudangrass 0.970.98 Fuchs and Tanner (1966)Rape 0.98 0.01 Rahkonen and Jokela (2003)Sow-thistle 0.98 0.01 Rahkonen and Jokela (2003)Mango 0.96 Arp and Phinney (1980)Pine 0.982 0.009 Arp and Phinney (1980)Olive 0.976 0.006 Rubio et al. (1997)Alfalfa 0.987 0.004 Rubio et al. (1997)Pine 0.982 0.009 Rubio et al. (1997)Holm Oak 0.985 0.010 Rubio et al. (1997)

Table 6Emissivity val upper side and underside of the nine crops analysed (for each group havebeen used ve

Crop Group of leaves 4 Maximum difference between leaves

UndersideTomato Pepper CucumberCourgetteAubergineMelon WatermelGreen beaRed bean

Upper sideTomatoPepper CucumberCourgetteAubergineMelon WatermelGreen beaRed bean

Within each li

melon (Fig. 0.004 and 0Finally, pepences betwand red bethe averagevalues are gsivity valuein part be dside. Red beis most not

Consideemissivity vcultural crowith a diffeFor future wthe temperof radiationshould be g

We havrecting for and undersature equal0.98 and rebe 0.74 C. Nicotiana tabacum L.Artemisia tridentata L. Medicago sativa L. Sorghum vulgare var. sudanense Hitchc. Brassicca rapa L. Sonchus arvensis L. Manginefara indica L.Pinus leiophylla Schlecht. and Cham.Olea europea L. Medicago sativa L. Pinus nigra Arnold. Quercus ilex L.

ues (average value standard deviation) obtained for different groups of leaves for leaves, one for each water bath temperature from 35 to 45 C).

Group of leaves 1 Group of leaves 2 Group of leaves 3

0.980 0.013a,b 0.981 0.009b 0.980 0.007b0.978 0.005a 0.978 0.008a,b 0.979 0.006b

0.983 0.003a 0.983 0.005a 0.983 0.007b0.985 0.003a,b 0.985 0.006a,b,c 0.985 0.007c

0.973 0.014a 0.973 0.013a 0.973 0.007a0.977 0.003a 0.978 0.006b 0.977 0.010a

on 0.981 0.005a 0.981 0.008a 0.982 0.014an 0.983 0.005a 0.983 0.005a 0.983 0.008a

0.983 0.004b 0.983 0.006b 0.983 0.008b0.981 0.010b 0.981 0.007b 0.980 0.008a0.987 0.006a,b 0.987 0.008b 0.986 0.006a0.986 0.004a 0.986 0.006b 0.985 0.005a

0.984 0.006c 0.985 0.009c 0.984 0.011b 0.968 0.008a 0.968 0.011a 0.968 0.005a

0.980 0.006a,b 0.980 0.006a 0.980 0.006a,bon 0.984 0.010a 0.985 0.008b 0.985 0.009a,bn 0.968 0.012a,b 0.968 0.015a,b 0.968 0.016b

0.954 0.007a,b 0.954 0.008a,b 0.954 0.012b

ne, the levels containing the same letter form a group of means within which there are n

6c) and aubergine, with differences in average values of.005 respectively, also less than the standard deviation.per, green bean and red bean showed greater differ-een the emissivity of opposite sides of the leaf. Greenan in particular (Fig. 6d), for which the differences in

values were 0.015 and 0.029, respectively, and thesereater than the standard deviation. The different emis-s observed between the opposite sides of the leaf mayue to the different tones of the upper side and under-an is one of the crops where this difference in tonalityable (Fig. 7).rable differences have been observed between thealues on the opposite sides of the leaves in some horti-ps, such as the green bean, and particularly the red bean,rence in the average emissivity value of 0.029 (Table 4).orks, in which researchers wish to evaluate or control

ature of horticultural crops carrying out measurements emission in the infrared range, special considerationiven to these crops.e estimated the possible error as a result of not cor-the difference in emissivity between the upper sideide of leaves of red bean. For instance, at leaf temper-

to 25 C measured for a recommended emissivity ofected temperature equal to 270 K, this error wouldLeaf temperature would be 24.93 C (emissivity 0.983)

and 25.66

leaves, respAny ima

include boters. This mdifcult to which sidetomato cro

Fig. 0.980 0.007a 0.00090.978 0.010a,b 0.00070.983 0.009a 0.00090.984 0.004a 0.00040.973 0.007a 0.00040.978 0.006b 0.00060.981 0.009a 0.00040.983 0.006a 0.00030.982 0.003a 0.00070.981 0.009b 0.00160.987 0.007a,b 0.00070.986 0.008a 0.00100.983 0.004a 0.00180.968 0.008a 0.00030.980 0.006b 0.00050.985 0.008b 0.00100.967 0.015a 0.00090.953 0.014a 0.0011

o statistically signicant differences (95% condence level).

C (emissivity 0.954) for the upper side and underside ofectively.ge taken of a real crop in the greenhouse is most likely toh upper sides of some leaves and the undersides of oth-akes later analysis extremely complicated as it is mostdifferentiate which areas of the image correspond to

of the leaf. Fig. 8 shows a thermographic image of ap taken inside a greenhouse. It would be impossible

8. Thermographic image of a tomato crop taken at midday.

A. Lpez et al. / Scientia Horticulturae 137 (2012) 4958 57

Fig. 9. Emissivhorticultural c

to tell aparreference p

3.6. Variati

In ordercrops, the vhave been aysis for the other.

The diffeof leaves arare all closeaubergine l

Greater for the undred bean, gbelow 0.98side of leave0.980 (melodeviation v

For the mwithin the of 0.980 is studied hertechnique tpurposes represent pap

4. Conclus

The metwithin a vevariation (