Leaf area index and optical properties measurements of raspberry in Hyytiälä, 2013
(+ some GNSS observations in Hyytiälä & Siikaneva)



Measurements conducted June 24 - July 22, 2013.
Last modified 22.7.2013

Note: All image files and raw data not available in the web version.

Photographs taken during the campaign (folder).


1. Specific leaf area measurements

All the samples collected from one 4-yr-old seedling stand (kansalaisen karttapaikka)
Leaf samples were weighed, scanned and dried in 80 °C oven.
Raspberry has leaves containing 3/5 leaflets, connected by a petiole, weighing separately for the leaflets and the petioles, in order to determine the proportion of the petiole from the leaf dry weight.
Scanned images thresholded at blue channel using Otsu's method, and the projected leaf area was determined from the thresholded image.

a. Samples #1
  Collected on 24.6.2013
  10 shoots in total, 5 1st year and 5 2nd year shoots.
  Each 1-yr/2-yr pair is from the same location.
  From each shoot, 3 samples of leaves were selected (from upper, middle and lower 1/3 of the leaf-containing part of the shoot).

b. Samples #2
  Additional samples collected on 27.6.2013.
  20 shoots in total, 10 1st year and 10 2nd year shoots, from different locations.
  One leaf sample from each shoot was collected, from the upper 2/3 of the leaf-containing part.
 
Measurement data for both samples (xlsx), see the sheet 'sla'.
Scanned images, samples #1 (folder)
Scanned images, samples #2 (folder)
Image thresholding, Matlab code (file)


2. Sample shoots for determining intershoot leaf area variation
  
Measurements conducted 25.-26.6.2013.

One 1×1 m area selected, same stand as for specific leaf area samples.
40 shoots from the area randomly selected (20/20, 1st/2nd yr).
Leaves from each shoot dried and weighed to determine the leaf area.
The leaf-containing part in the shoot was divided into 3 sections of equal length, and the leaf area determined for each section separately.

Results in the excel file (xlsx), see the sheet 'shoot_leaf_area'.


3. LAI measurements from sample areas

Measurements conducted 28.6.-4.7.2013.

Circular areas (50) of homogeneous raspberry canopies (minimum LAI and height variation within the area) were marked in the field.
Each area is defined by its center, and a radius (65-100 cm).
A rectangular plot (1×1 m) was established at the center of each area.
Shoots collected from each plot, and brought to the laboratory for analysis.
Two types of measurements were conducted on the plots.

Type 1 measurements (21 plots):
  - visual canopy cover estimate (0-100 %)
  - maximum height
  - weighing all shoots (fresh weight, 1st year and 2nd year shoots separately)
  - selecting a sample of shoots for leaf area determination, separate samples for 1st year and 2nd year shoots
  - sampling ratio dependent on the total number of shoots
      <60 shoots --> 2/3
      60-90 shoots --> 1/2
      >90 shoots --> 1/3
  - aimed at 30-45 sample shoots (<10% standard error of LAI, based on simulations)
  - leaves were picked from the sample shoots and weighed
      fresh weight --> oven drying (80 °C) --> dry weight
  - leaf area for the plot calculated using dry weight of the leaves, and the sample ratio
  - the fresh/dry weight of leaves includes the petiole that connects the leaflets, its mass was reduced from the measured value using regression models from 1)
  - height measured for each sample shoot

Type 2 measurements (29 plots):
  - visual canopy cover estimate (0-100 %)
  - maximum height
  - weighing all shoots (fresh weight, 1st year and 2nd year shoots separately)
  - LAI predicted with regression models, using the fresh weights of 1st year and 2nd year shoots

The center of each sample plot was positioned with GNSS on 15.7. 2013. In addition, the plot corners were positioned on the type 1 plots.

Results in the excel file (xlsx) (LAI measurements: sheet 'LAI_data'; heights of the sample shoots: 'shoot heights'; GNSS observations: 'GNSS_obs').
Sampling simulator (R-code).
Photographs of the plots 1-14 (folder)


4. Leaf optical properties measurements

Measurements conducted 19.-21.7.2013.
Optical and structural properties of raspberry leaves (and a small sample of fireweed).
 
4.1 Sample locations

Four (4) locations of raspberry, and three (3) of fireweed were selected. The stand was the same as used for collecting specific leaf area samples in 1).
Selected fresh-looking, green canopies. The leaves were already yellowish in most parts of the stand due to the dry season.
Shoots (1-3) from each location were cut and brought to the laboratory. For raspberry, separate samples of 1st year and 2nd year shoots were selected.
While waiting for the measurements, the shoots were kept in a cooling room (7 ºC), in a bucket of water, the lights on.
The maximum time between cutting the shoots and making the measurement was ~4 hours.
Coordinates of the sample points measured with GNSS (csv).


Raspberry 1
Height 1.55 m, canopy cover (visual esimate) 80%
Next to the road, slope towards NW, open area.

Raspberry 2
Height 1.20 m, canopy cover 85%
Fertile site, alongside a ditch, open area.

Raspberry 3
Height 1.1 m, canopy cover 75%
Next to a mature forest stand, trees partly shading from the S side.

Raspberry 4
Height 1.3 m, canopy cover 75%
Next to a mature forest, trees on the S side, 7 m from the sample point.

Fireweed 1
Height ~1.5 m, canopy cover 80%
Open area, dense canopy.

Fireweed 2
Height ~1.5 m, canopy cover 80%
Open area, dense canopy.

Fireweed 1
Height ~1.5 m, canopy cover 70%
Open area, less dense.


4.2 Optical properties

Measured directional-hemispherical reflectance and transmittance factors (DHRF/DHTF) at 350-2500 nm, using ASD FieldSpec 3 PRO spectroradiometer and ASD RTS-3ZC integrating sphere.

Notation for naming the samples:
  - sample location
  - year class (1 = 1st year, 2 = second year; for raspberry only)
  - canopy position (E = sun exposed, S = shaded)
  Example:
  Raspberry: 1.2.E = location 1, 2nd year shoot, sun exposed
  Fireweed: 1.S = location 1, shaded

One fresh-looking leaf representing each sample was selected, detached from the shoot and measured right after the detachment.
4 measurements per sample were taken (reflectance and transmittance, both sides of the leaf).

Work order in the measurements was following:
  - putting on the spectrometer, heating the lamp (~30 min before the measurements)
  - optimization of the spectrometer i.e. the integration time is adjusted so that the whole DN range (16-bit) is used<>
  - measuring the samples:
    - for each measurement, 30 readings were taken and averaged
    - each measurement stored in a separate file, see the measurement log for file names
  - transmittance measurement was repeated 5 times for each sample, by detaching and attaching the leaf to the sample port at each time (this was done because variation was observed between transmittance measurements)<>
  - calculation of the reflectance/transmittance factors: DHRF/DHTF = (measured DN - nearest stray light) / (nearest white reference - nearest stray light)
  - stray light was measured at the beginning and at the end of measurements
  - white reference (separately for each measurement mode) was measured for every other sample

In addition to the basic samples, effect of moisture on raspberry leaf reflectance was tested by spraying the leaf with water.

Measurement log (xlsx).
Port configuration of the integrating sphere and the measurement modes (xlsx) (illustration).
Matlab code for processing the samples (file).
Reflectance spectra of raspberry (fig), transmittance spectra of raspberry (fig), reflectance and transmittance spectra of fireweed (fig).
Effect of moisture on raspberry leaf reflectance (fig)
Measurement data (ASD raw files and exported ASCII). ASD raw files can be readed with ViewSpecPro software (download).
Calculated spectra for the samples (xlsx), calculated after the measurements, no smoothing or cleaning of the outliers. Especially in the transmittance data should be studied more carefully.

4.3 Specific leaf area

Specific leaf area was measured for the same sample shoots as for optical properties measurements. Same notation in sample naming used.

Weighing of the leaves, scanning, drying in 80 °C oven for at least 24 hours.
Scanned images thresholded at blue channel using Otsu's method, and the projected leaf area was determined from the thresholded image.

Specific leaf areas for the samples (xlsx).


5. Other GNSS observations from Hyytiälä

VRS-GPS (Trimble R8) used.
Data in excel file (xlsx)
Raw data, can be read with Trimble Business Center software (folder).

- samples n:o 144-318: Point samples of sand road surface near the raspberry stands, for radiometric calibration of LiDAR signals. Samples contain pure sand/gravel, and no grass. The radius of each sample in centimeters (30-50 cm) is recorded. The LiDAR pulses inside the radius are safe to use for radiometric tests.


- samples n:o 319-1040: Point samples for evaluating z-accuracy of LiDAR data. The points are selected so that the surface is as level and smooth as possible within 50 cm radius from the sample center. If the radius is smaller, this is mentioned in the radius-column.
319-482: asphalt road
483-545: sand road
546-738: sand road, may contain some grass
739-765: sand road
766-887: sand road, may contain some grass (Note! When compared against Z data from LiDAR, the points 782-808 have 0.6 m bias. They are clearly erroneous, DO NOT USE THOSE POINTS!. Reason is unknown. Probably the poor visibility of satellites.)
888-1040: asphalt road

- samples 1041-1145, 1231-1400, 1501-1525: Samples of fireweed and birch
Each sample has attributes: radius [cm] (measured for fireweed, 0 for birch samples), species [1=birch, 2=fireweed], visual canopy cover estimate [%], height [cm]
Fireweed has almost certainly gained height after the 16.6. LiDAR campaign. Example pictures from the burned stand near the SMEAR station

In total, 215 birches and 160 m2 (85 samples) of fireweed measured.
:
- samples 1146-1167, 1171-1230, 1401-1500: Sample points for evaluating the accuracy of the GPS equipment
Sample point 1 (Pilkotun tie):
Trees blocking the view, measured using tolerance of 0.03 m in horizontal, 0.05 m in vertical direction
Measured on 16.6. (1146-1167), repeated on 17.7. (1401-1432)
Sample point 2 (Pilkotun tie):
Trees blocking the view, measured using tolerance of 0.03 m in horizontal, 0.05 m in vertical direction
Measured on 16.6. (1171-1200), repeated on 17.7. (1433-1469)
Sample point 3 (Hyytiälä station):
Open area, good accuracy, tolerance 0.015 m in horizontal, 0.03 m in vertical direction
Measured on 16.6. (1201-1230), repeated on 17.7. (1470-1500)


6. GNSS-mittauksia Siikanevan ilmakuvien ja laserkeilauksen (28.5.2013) laadun tarkistukseen

Mittaukset tehty 17.7.2013 Siikanevan mittaustornien (Siikaneva 1 ja 2) ympäristöstä VRS-GPS laitteistolla (Trimble R8).
Pisteet excel-tiedostossa (xlsx).

- pisteet 1-21 (Siikaneva 1), 83-102 (Siikaneva 2): Pitkospuiden ja paalujen kulmia ilmakuvien orientoinnin tarkastamiseen
1: paalun kulma
2-21: pitkospuun kulmia
83-102: pitkospuun kulmia
Piirrokset pisteistä (1-10, 11-18, 19-21, 83-91, 92-99, 100-102)
- pisteet 22-82 (Siikaneva 1), 123-172 (Siikaneva 2): Suon pintapisteitä laserkeilauksen z-tarkkuuden arviointiin.
Valittu pisteitä, joissa pintakasvillisuus vähäinen ja maanpinta mahdollisimman tasainen 20 cm säteellä pisteestä. Esimerkkikuvat alla.

GPS-antennia kannateltu mitattaessa kevyesti: antennitangon alaosa koskettaa maanpintaa, mutta antenni ei uppoa turpeeseen.
Siikaneva 1:llä mittaukset tehty pitkospuiden vierestä, pitkospuilla seisten. Siikaneva 2:lla mittauksia myös pidemmältä pitkospuiden ulkopuolelta.
- pisteet 103-122: Testimittauksia yhdestä mittauspisteestä Siikaneva 1:lla: selvitetty mittausten hajontaa.