spectral sensing of foliar water conditions in conifer needles

Spectral sensing of foliar Spectral sensing of foliar water conditions in conifer water conditions in conifer

needlesneedles

Hugh StimsonHugh Stimson11, David Breshears, David Breshears22, , Shawn KefauverShawn Kefauver11, Susan Ustin, Susan Ustin11

11 University of California, Davis University of California, Davis2 2 Los Alamos National LaboratoryLos Alamos National Laboratory

Ecological Society of America Annual Meeting, Savannah Georgia USA, August 7Ecological Society of America Annual Meeting, Savannah Georgia USA, August 7 thth 2003 2003

goalsgoals

•find relationships find relationships between water and between water and reflectance at the reflectance at the needle levelneedle level

•assess a suite of assess a suite of reflectance indicesreflectance indices

•lay groundwork for lay groundwork for pending aerial pending aerial imaging of water imaging of water statusstatus

systemsystem

protocolprotocolsample gradientsample gradient

• 17 trees selected17 trees selected

• selected from visual selected from visual estimate for wide water estimate for wide water rangerange

water contentwater content

• 3 twigs per sample3 twigs per sample

• 11stst whorl of needles removed whorl of needles removed

•weighed, dried to constant weight, weighed, dried to constant weight, reweighedreweighed

water potentialwater potential

• N = 15N = 15

• average of 2 twigs per treeaverage of 2 twigs per tree

• measured in pressure bombmeasured in pressure bomb

protocolprotocol

spectral reflectancespectral reflectance

• same 3 twigs as water same 3 twigs as water contentcontent

• subset of needles subset of needles removed, racked onto vinyl removed, racked onto vinyl mountsmounts

• measured with ASD full-measured with ASD full-range field spectrometer range field spectrometer via Licor integrating spherevia Licor integrating sphere

• Measured within 15 Measured within 15 minutes of fridge, 5 of minutes of fridge, 5 of instrument calibrationinstrument calibration

•integration time 10sintegration time 10s

protocolprotocol

protocolprotocol

datadata

datadata

datadata

datadata

analysisanalysis

direct water direct water

measurementsmeasurementsv.v. spectral indicesspectral indices

Water contentWater content

(% foliar mass)(% foliar mass)

Water potentialWater potential

(MPa)(MPa)

980nm water absorption

1200nm water absorption

normalized differential water index

normalized differential vegetation index

analysisanalysis

analysisanalysis

970nm

analysisanalysis

1200nm

analysisanalysis

NDWI860-1240

860+1240

analysisanalysis

NDVI880-660

880+660

analysisanalysis

NDWI

970nm1200nm

NDVI

resultsresults

foliar water content (%mass)foliar water content (%mass)

997700

nnm m

RR22=0.91=0.91

p<0.001p<0.001

0% 25% 50% 75%

spectral indices v. water contentspectral indices v. water content

11220000

nnm m

RR22=0.59=0.59

p<0.001p<0.001

NNDDWWII

0% 25% 50% 75%

RR22=0.96=0.96

p<0.001p<0.001

0% 25% 50% 75%

RR22=0.93=0.93

p<0.001p<0.001

NNDDVVII

0% 25% 50% 75%

resultsresults

water potential (|MPa|)water potential (|MPa|)

997700

nnm m

RR22=0.37=0.37

p=0.03p=0.03

spectral indices v. water potentialspectral indices v. water potential

11220000

nnm m

RR22=0.03=0.03

p=0.5p=0.5

NNDDWWII

RR22=0.45=0.45

p=0.006p=0.006

RR22=0.35=0.35

p=0.02p=0.02

NNDDVVII

1.5 2.5 3.5

1.5 2.5 3.5

1.5 2.5 3.5

1.5 2.5 3.5

resultsresults

Water Content Water Potential

R2 p R2 p

#1 NDWI 0.93 <0.00 0.45 0.01

#2 NDVI 0.93 <0.00 0.35 0.02

#3 970nm 0.91 <0.00 0.37 0.03

#4 1200nm 0.59 <0.00 0.03 0.50

resultsresults

•The foliar water content of Piñon pine needles The foliar water content of Piñon pine needles can be reliably predicted from their reflectance.can be reliably predicted from their reflectance.

• A weaker, indirect, relationship exists A weaker, indirect, relationship exists between water potential and reflectance.between water potential and reflectance.

• The most robust measure is NDWI, closely The most robust measure is NDWI, closely followed by NDVI and 970nm absorption.followed by NDVI and 970nm absorption.