Vegetation indices calculated from reflectance spectra collected at Arctic LTER experimental plots at Toolik Lake, Alaska during the 2007-2019 growing seasons.
Long term experimental plots span several different vegetation types: Heath (HTH89), Moist Acidic Tussock (MAT89 and Low Fert), Moist Non-Acidic Tussock (MNAT), Non-Acidic Non-Tussock (NANT), Shrub (SHB), and Wet Sedge (WSG). Plots are differentiated by their experimental treatment and are located in replicate blocks.Canopy reflectance is measured by hand-held spectrophotometer and several indices of interest (NDVI, EVI, EVI2, PRI, WBI, and Chlorophyll index) are calculated.
Data set ID:
EML revision ID:
EXPERIMENTAL DESIGN AND METHODS
Long term experimental plots spanning several different vegetation types have been established at Toolik Lake, Alaska. Vegetation types include Heath (HTH), Moist Acidic Tussock (MAT and Low Fert), Moist Non-Acidic Tussock (MNAT), Non-Acidic Non-Tussock (NANT), Shrub (SHB), and Wet Sedge (WSG). Plots are differentiated by their experimental treatment and are located in replicate blocks.
Spectral monitoring of LTER experimental plots began in 2007 and has continued through 2010. Canopy reflectance is measured by hand-held spectrophotometer and several indices of interest (NDVI, EVI, EVI2, PRI, WBI, and Chlorophyll index) are calculated.
In 2007 and 2008, a single channel spectrophotometer (Unispec SC,PP Systems,Haverhill, Massachusetts, USA) was utilized. A specific sampling procedure was not outlined so comparison to later data should be done cautiously.
In 2009, 5m transects along edges of LTER plots were chosen for spectral monitoring. Diagrams of these designated areas are available.
In 2009 and 2010, the plot transects were scanned for canopy reflectance several times throughout the growing season. For each plot, a dual channel spectrophotometer (Unispec DC, PP Systems,Haverhill, Massachusetts, USA) was used to make five replicate scans located 1 meter apart along the designated 5-meter transect. Scans were taken from a height of 1.09m approximately 0.5 meters from the edge of the plot and have a circular field of view with a radius of approximately 40 cm.
In 2015 the protocol from 2009 and 2010 was followed at all sites.
Canopy reflectance was measured using a dual channel spectrophotometer (Unispec DC, PP Systems, Amesbury, Massachusetts, USA) in 2009-2010 and 2015. This instrument utilizes two foreoptic cables to simultaneously receive incoming irradiance and radiance reflected by the canopy vegetation. One foreoptic cable (UNI-684) is oriented downwards and is equipped with a ferrule over which a 100mm hypotube (UNI-688) is placed. This produces a field of view that extends at an angle of 20 degrees from the end of the hypotube. The other foreoptic cable (UNI-686) is oriented upwards and is fitted with a cosine receptor (UNI-435). The two foreoptic cables are connected to two miniature photodiode array detectors that produce signals ranging from zero to 65,000 A/D counts for 256 wavebands. These wavebands represent 3.3 nm wide portions of the visible and near infrared spectrum from 310 to 1100 nm. A scan is performed over a period of milliseconds with the exact integration time determined by the user based on current light conditions.
At the time of data collection, a dark scan is performed by covering the foreoptics with a dark cloth. The raw signals from the dark scan are used by the machine to correct for background noise. A reference scan is also performed by positioning the downward foreoptic cable over a white standard (UNI-420). Canopy reflectance is calculated for each waveband as follows:
Reflectance= (Idata down / Idata up) x (Ireference up/ Ireference down)
Idata down = signal from downward foreoptic during data scan (radiance reflected from target vegetation)
Idata up = signal from the upward foreoptic during data scan (incoming irradiance)
Ireference up= signal from the upward foreoptic during reflectance scan (incoming irradiance)
Ireference down= signal from the downward foreoptic during reflectance scan (radiance reflected from white standard)
Data is processed and interpolated from 3.3nm wavebands to single nanometer intervals using the program Multispec V.5 (available at http://specnet.info/specnet_toolkit.htm).
Spectral indices were calculated for each reflectance spectra as follows:
NDVI (MODIS) = (NIR-Red)/(NIR+Red)
EVI (MODIS) =2.5*(NIR-Red)/(NIR+6*Red-7.5*Blue+1)
EVI2 (MODIS) =2.5*((NIR-Red)/(NIR+2.4*Red+1))
PRI (550 Reference) =(550nm-531nm)/(550nm+531nm)
PRI (570 Ref) = (570nm -531nm)/(570nm+531nm)
Chl Index =(750nm-705nm)/(750nm+705nm)
For MODIS indices, Spectral bands are defined as follows:
Reflectance values equal to or less than zero were replaced with "NaN" before indices were calculated.
Reflectance values equal to or greater than one were replaced with "NaN" before indices were calculated.
Vegetation indices were calculated for replicate scans and then averaged per block and collection day to produce the data and standard deviations presented here.
Start of experimental treatments:
HTH( 1989), MAT (1989 except exclosure plots which were started in 1996), MNAT (1997, except sulfur and lime treatments which were started in 1998), NANT (1997), SHB (1988), WSG (1989), 06MAT also known as the low fert MAT site started in 2006
Experimental treatments are as follows:
ct-control, no fertilization.
ct2- control, differentiation by number when more than one control plot exisits per block.
ct3- control, differentiation by number when more than one control plot exisits per block.
exctsf-Herbivore exclosure, control, small fence.
exctlf-Herbivore exclosure, control , large fence.
exnpsf-Herbivore exclosure, NP fertilization, small fence.
exnplf-Herbivore exclusure, NP fertilization, large fence.
f0.5- Fertilized with 0.5g N + 0.25g P per square meter. N is in form of NH4NO3.
f1- Fertilized with 1g N + 0.5g P per square meter. N is in form of NH4NO3.
f2- Fertilized with 2 g of N + 1g P per square meter. N is in form of NH4NO3.
f5- Fertilized with 5g of N + 2.5g P per square meter. N is in form of NH4NO3.
h0- Reserved for Herbivory project (Moore and Gough)
nh4-Fertilized with 5g of N + 2.5g P per square meter. N is in form of NH4Cl.
lime- Fertilized with 150g lime per square meter.
n -Fertilized with 10g N per square meter. N is in form of NH4NO3.
NP- Fertilized with 10g N + 5g P per square meter. N is in form of NH4NO3.
no3-Fertilized with 5g of N + 2.5g P per square meter. N is in form of NaNO3.
p- Fertilized with 5 gP per square meter.
s- Fertilized with 50g sulfur per square meter.
x0- Extra plots, as yet no treatment.
x1-Extra plots, as yet no treatment.
x2-Extra plots, as yet no treatment.
OTHER DATA FILES TO REFERENCE:
2007-2016gsreflectancespectra (full reflectance spectra for all scans)
Full Metadata and data files (either comma delimited (csv) or Excel) - Environmental Data Initiative repository.
Use of the data requires acceptance of the data use policy --> Arctic LTER Data Use Policy