Santa Monica Mountains Field Trip (June 8-12 1995)

Susan Ustin, Claudia M. Castaneda, Stephane Jacquemoud, George Scheer
Department of Land, Air and Water Resources
University of California, Davis, CA 95616
Address for Correspondence:
Dr. Susan L. Ustin
Department of Land, Air, and Water Resources
University of California
Davis, CA 95616
Phone: (530) 752-0621
FAX: (530) 752-5262
email: slustin@ucdavis.edu

This document is a report describing the events that took place during June 8 through June 12, 1995 for the Santa Monica Mountains experiment.

1.  Introduction

The experiment was to obtain field canopy reflectance measurements and fresh weight canopy measurements to provide enough information to calibrate AVIRIS images to surface reflectance. Three sites, Zuma Ridge, Castro Crest and Encino Reservoir, were chosen as representative of the vegetation under study and presenting vegetation in different stages of growth.

2.  Description of Sites and Vegetation

Zuma Ridge: a coastal site with young and mixed vegetation
Castro Crest: mountain site with medium growth and mixed vegetation
Encino Reservoir: inland site on the side of the reservoir with old growth and one type of vegetation, chaparral 3 to 4 meters high.
The following are descriptions of the vegetation studied; first we present the site and the species under study at each of them; Table 1 presents a summary of the species giving their scientific and common names and their family, finally a general description of each plant is given.
 
Zuma: MALA Castro: ADFA Encino: CEME
ARCA CEOL DRY GRASS
SALE ARGL
ERAR
 
Acronym Latin name Family  Common name
MALA Rhus laurina Anacardiaceae Laurel Sumac
ARCA Artemisia californica Asteraceae Coastal Sagebrush
SALE Salvia leucophylla Lamiaceae Purple Sage
ERAR Eriogonum cinereum Polygonaceae Ashy Leaf Buckwheat
ADFA  Adenostoma fasciculatum Rosaceae Chamise, Greasewood
CEOL Ceanothus oliganthus Rhamnaceae Hairy-leaf Ceanothus 
ARGL Arctostaphylos glandulosa Ericaceae Eastwood Manzanita
CEME Ceanothus megacarpus Rhamnaceae Big Pod Ceanothus
Table 1.  List of species found in the three sites.

Laurel Sumac

Laurel Sumac is an evergreen shrub with smooth, reddish-brown bark that grows 6 to 15 feet tall. The oblong leaves which might be 4 inches long are untoothed on the margin, slightly folded along the midrib, have an abrupt sharp tip and smell like bitter almonds when crushed. The tiny, white flowers which are enjoyed by the bees cluster densely at the ends of the branches. The fruit is smooth and white and yields an oil. This is a common shrub throughout our mountains and blooms in June and July.
 

Coastal Sagebrush

Coastal Sagebrush is a much-branched shrub 2 to 5 feet tall. The numerous grayish-green leaves are once or twice-parted into thread-like divisions. The greenish flower heads, composed of disk florets only, are very small, nodding on tiny stalks crowded along 4 to 12 inches of the terminal stems. This is one of the dominant plants in Coastal Sage and occurs throughout our mountains usually below 2000 feet. It blooms from August to February. The leaves of Coastal Sagebrush have a clean, bitter, pleasantly aromatic fragrance.
 

Purple Sage

The Purple Sage is a shrub 3 to 5 feet tall with gray, hairy herbage. The leaves, 3/4 to 2 inches long, are longer than wide and have small rounded teeth on their edges. They are heart-shaped at the base and have prominent veins. The rosy-lavender flowers are in compact whorls with gray, oval, leaf-like bracts underneath. Purple Sage is common in coastal Sage throughout the area blooming from May to July. Both the stems and the bracts below the flowers feel mealy to the touch.
 

Ashy Leaf Buckwheat

Ashy Leaf Buckwheat, 2 to 3 feet high,  has oval leaves and a gray appearance and flowers in the months of April to October. It is most abundant near the coast.
 

Chamise, Greasewood

The Chamise is a shrub 2 to 15 feet high with evergreen, needle-like, 1/4 inch leaves in clusters along the branches. The white flowers, individually very small, are disposed in showy, crowded, compound clusters several inches in height and terminal on the branches. The fruit is an achene. Chamise is one of the dominant plants found throughout Chaparral and Coastal Sage away from the immediate coast. It blooms from April to June, In the Coast Ranges in early spring, Chamise often covers miles of slopes with first a dense uniform green. Then in April and May, the slopes are diffused in snowy white blooms followed by a warm bronze from the abundant seed vessels. Greasewood is a name given to many southwestern plants. In this case, the dry branches are most flammable as though they do indeed contain grease. They often contribute to the spread of brush fires.
 

Hairy-leaf Ceanothus

The small leaf-like structures called stipules at the base of the leaf or leaf stems are thin and fall early. The leaves are smooth and the fruit is not horned. 3-veined leaf with a toothed edge and covered with soft short hairs on the top; blue flowers; flexible branches; blooms March and April; especially abundant on Castro Peak.
 

Eastwood Manzanita

This is a spreading shrub 2 to 4 feet high with several smooth, reddish, crooked stems from a basal burl which crown sprouts after fire. The oval or lance-shaped leaves, over an inch long and often an inch wide, are light to dark green, softly hairy on both surfaces and sharply pointed. The white flowers, 1/2 inch long, are urn-shaped. The red-brown fruit is round and sticky. Eastwood Manzanita may be found in open Chaparral away from the coastal throughout flowering from January to March.
 

Big Pod Ceanothus

The stipules are thick and hang on; the leaves are rough and have tiny holes (stomata) on the underside; horns are on the fruit; the flowers are always white; leaves alternate; often composes 50 percent of the cover of southern slopes; blooms January to April.

References: N. Dale (1986), Flowering plants - The Santa Monica Mountains, Coastal & Chaparral Regions of Southern California, Capra Press (Santa Barbara), 239 pp.

3.  Measurements taken on Vegetation

3.1.  Field Radiometric Data

The procedure followed to obtain calibrated canopy spectra data is described in the next paragraphs.

1.   For all three sites, seven flag locations were chosen from the bucket above the canopy for the radiometric measurements; at the Zuma and Castro sites, the flags, made from red and white palstic flagging tape, were placed from the ground in such a way that they could easily be seen from the bucket at the measured height, at Encino the flags locations were only accessible via the bucket. Figure 1 represents the approximate location of the flags.

  Figure 1
 Figure 1. Azimuth orientation of the bucket and detector.

After choosing and flagging the sites, the bucket was stabilized vertically for height measurements. The bucket was swung out along a horizontal arc, keeping the height constant with respect to the ground. The varying height between the detector and the crown of the canopy was subsequently measured for each flag location and recorded by someone on the ground. Table 2 presents the height over the ground of the detector, the specie and the flag number assigned to it.
 
 
Flag
Zuma
Castro
Encino
1
 MALA 3.4 m ADFA 3.1 m CEME 2.8 m
2
 ARCA 3.4 m ADFA / ARGL 3.4 m CEME 3.2 m
3
 SALE 3.0 m CEOL 3.3 m CEME 2.5 m
4
 SALE / ARCA 3.5 m CEOL 3.3 m CEME 2.5 m 
5
 ARCA 3.6 m CEOL / ADFA 3.4 m CEME 2.6 m
6
 ERAR / ARCA 4.5 m ADFA 3.2 m CEME 3.5 m
7
 SALE 4.0 m ADFA 3.6 m GRASS 6.0 m

Table 2. Heights of detector above canopy (in meter) and corresponding species

2.  The Spectralon standard was mounted on a tripod attached to the bucket and adjusted normal to the ground using a leveling device taped to the corner of the standard.

3.  After preparing the flag sites, the ASD and laptop were powered-on and the optimization process was begun. The new ASD model (the "Full Range" model) uses three detectors, each requiring an optimized integration time. The process required that the gun-detector was pointed downwards toward the white standard (the 99% reflectance portion of the standard) until the spectrometer found the three best integration times for each of the detectors. Usually the process required about a minute and, at times, required a second attempt if one of the detectors failed to optimize (the third detector in the NIR region usually failed the first attempt).
 
4.   Once the three detectors were optimized, the canopy reflectance measurements were made by positioning the detector gun as closely to the original height position as possible (an arms length) and pointing the detector gun as nadir as possible to the flagged locations. For the first measurement, a 99% reflectance standard was scanned followed by five measurements of the canopy, then ending with another 99% reflectance standard. The recorder on the ground and the bucket-rider continually verified the consistency between the ASD filenames and the written record of the filenames and descriptions. A compilation of all this data can be seen in the annex to this document

5.   Once the first flagged site was successfully scanned, the bucket was moved horizontally to the next flagged location, keeping the original vertical position and the same procedure for acquiring canopy spectra was used: first a white standard, then five replicates of the flagged site, then another white standard while at the same time verifying with the recorder the filename numbers and descriptions of each number.

6.   Information about the measurements taken from each site can be found in Tables 3.1, 3.2, and 3.3. The files containing the raw data are named:

    Canopy_Spectrum_Year_Month_Day_Pass.FileNumber

For instance, \zuma\cs950608a.000 corresponds to the white standard reflectance acquired on June 8th in the Zuma site for the first pass (see Table 3.1.).

7.   Figure 2 shows the raw signal measured by the ASD on the standard (two spectra xs1 and xs2) and on vegetation canopy (five spectra xc1 to xc5), as well as the absolute reflectance of the standard (js). We first averaged the standard and canopy spectra to obtain xs and xc. The calibrated canopy reflectance spectrum jc is calculated using:

 The Matlab routine canspec.m and san_mon.m was used to correct at one go all the spectra. The new spectra can be in found in zuma.dat, castro.dat, and encino.dat. The first column is the wavelength ranging from 400 to 2500 nm with a 2 nm step; the other columns contain the canopy reflectance spectra. The relation between the spectrum position in the file and the actual target measured can be found in Tables 3.1 to 3.3. Figures 3 and 4 illustrate the reflectance spectral and directional variation measured in the Zuma site.
 

Figure 2
Figure 2. Raw signal measured by the ASD on the standard (E_panel) and on vegetation (E_canopy),
as well as the absolute reflectance of the panel (R_Halon).
Figure 3
Figure 3. Calibrated reflectance of seven vegetation canopies in the Zuma site.
 
 Figure 4
Figure 4. Directional variation of canopy reflectance during the day.
 

3.2.  Laboratory Radiometric Data

For most of the species picked from all three sites, both leaf reflectance and transmittance were measured on the CARY 5E spectrophotometer. We acquired reflectance spectra for all the species. For ADFA and ARCA characterized by needle-like leaves, the transmittance could not be measured so that only the infinite reflectance of an optically thick sample was available. In short, the leaves from each species were processed for measurement according to their size and shape. The port hole on the CARY 5E was approximately 3 cm in diameter and the light path was 1 cm by 2 cm (for reflectance). Below is a summary of the procedure used to prepare the samples for measurement:

ARGL (Arctostaphylos glandulosa): two leaves were cut at 1/5 their original width and combined along their edges to ensure that all the light touched the leaves.
 
CEOL (Ceanothus oliganthus): three to four leaves were cut and combined in a similar manner to ARGL. These leaves were smaller so more were required to completely cover the light path.

 
ADFA (Adenostoma fasciculatum): these needle-like leaves were packed into an optically sealed container with a diameter of about 4 cm and a depth of about 2 cm, and were measured for reflectance only. About 2 cm thick of sample was placed into the container for measurement.

CEME (Ceanothus megacarpus): these leaves were small like CEOL and cut in a similar manner using 4 or 5 leaves.

ERAR (Eriogonum cinereum): these leaves were small like CEOL and cut in a similar manner using 4 or 5 leaves.

SALE (Salvia leucophylla): these leaves were typically larger, enough to completely fill the light path of the CARY 5E. Some flowers were put into the optically sealed container and only scanned for reflection.

MALA (Rhus laurina): these leaves were typically larger, enough to completely fill the light path of the CARY 5E.

ARCA (Artemisia californica): we followed the same procedure for these needle-like leaves as for ADFA.

The leaf reflectance spectra had to be calibrated. First, the reflectance of single leaves was measured using a black card of reflectance (  ) as a background. The non-zero reflectance of this black card induces an overestimation of the leaf reflectance. Assuming that  (  ) and  (  ) are respectively the leaf reflectance and transmittance measured on the same blade, the true leaf reflectance R(  ) can be calculated by:

 

Corrections for Spectralon were also post-processed to produce absolute 100% reflectance values regardless of the white standard employed. The previous leaf reflectance R have been multiplied by the known reflectance of the reference material to obtain the actual reflectance of the samples. The Matlab leafspec.m routine was used for this purpose. We gathered all the spectra in the file leaf.dat which contains 100 columns: the first one is the wavelength, the other ones are reflectance (# j) and transmittance (# t); spectra has been arranged as described in Table 4. For optically thick samples (needle-like leaves and flowers), three infinite reflectance spectra are available. The wavelengths range from 400 nm to 2500 nm with an interval of 2 nm.
 
Zuma # j # t # ji Castro # j # t # ji Encino # j # t
erar32 leaf 2 3   argl01/06 leaf 38 39   ceme20 leaf 62 63
erar33   4 5   argl02/07   40 41   ceme21   64 65
erar34   6 7   argl03/08   42 43   ceme22   66 67
erar35   8 9   argl04/09   44 45   ceme23   68 69
erar36   10 11   argl05/10   46 47   ceme24   70 71
erar37   12 13   argl11/11   48 49   ceme25   72 73
sale38 leaf 14 15   ceol12 leaf 50 51   ceme26   74 75
sale39   16 17   ceol13   52 53   ceme27   76 77
sale40   18 19   ceol14   54 55   ceme28   78 79
sale41   20 21   ceol15   56 57   ceme29   80 81
sale42   22 23   ceol16   58 59   ceme30   82 83
sale43   24 25   ceol17   60 61   ceme31   84 85
sale44 flower 86 87 88 adfa18 needle 95 96 97        
mala45 leaf 26 27   adfa19   98 99 100        
mala46   28 29                    
mala47   30 31                    
mala48   32 33                    
mala49   34 35                    
mala50   36 37                    
arca51 needle 89 90 91                  
arca52   92 93 94                  

Table 4. Leaf optical properties measured in the laboratory.

Figure 5a and 5b gather a few spectra of vegetation material collected in the Zuma site.
 

Figure 5a
Figure 5b
Figure 5. Leaf optical properties of the few samples collected in the Zuma site: reflectance and transmittance spectra of Laurel Sumac (Rhus laurina) [left] reflectance spectra of ERAR leaf (e), SALE leaf (s) and flower (s_f), MALA leaf (m), and ARCA leaf (a) [right].

 3.3.  Laboratory biophysical Measurements

 Some samples of fresh leaves, stems and flowers were collected in the field to calculate water content variations throughout the day. For large plant leaves, the fresh weight of 3.46 cm2 disks taken using a cork borer was immediately measured; for small leaves, we weighted entire blades, the area of which has been later measured using a camera and a digitizer. The stems and flowers of some plants were also processed. All the samples were placed into paper bags marked with a pre-established nomenclature and then placed in a drying oven at 70ºC for four days before being re-weighed. Assuming that FW is the fresh weight, DW the dry weight, and S the leaf area, when available, we calculated the water content (WC), the equivalent water thickness (EWT), the leaf specific weight (LSW) and the specific leaf area (SLA) which is the reciprocal of the leaf specific weight:

 WC = (FW-DW)/FW         EWT = (FW-DW)/S         LSW = 1/SLA = DW/S

WC is the water mass over fresh mass, EWT and LSW are respectively the water and dry matter masses per unit leaf area, expressed in g.cm-2; in consequence, the SLA is provided in cm2.g-1. These measurements were repeated as plant materials were scanned in the laboratory. Detailed results can be found in the Matlab file water.m but Table 5 gathers average values for each plant species. Some leaf samples have been frozen for later pigment concentration measurements.
 
Species Site Plant Material WC EWT LSW SLA
MALA Zuma pm/leaf 0.6286 0.0309 0.0183 54.64
Castro noon/leaf 0.6268 0.0253 0.0147 69.03
  pm/leaf 0.6390 0.0251 0.0142 70.42
  noon/stem 0.6740      
  pm/stem 0.7037      
Spectro leaf 0.5000 0.0298 0.0298 33.56
ARCA Zuma pm/leaf+stem 0.6892      
Spectro leaf 0.6239      
SALE Zuma pm/leaf 0.7017 0.0235 0.0100 100.0
  pm/stem 0.7114      
  pm/flower 0.7517      
Spectro leaf 0.6036 0.0192 0.0126 79.37
  flower 0.7030      
ERAR Zuma pm/leaf 0.6796 0.0217 0.0102 98.04
Spectro leaf 0.5507 0.0176 0.0142 70.42
ADFA Castro noon leaf+stem 0.4477      
  pm/leaf+stem 0.4264      
Spectro leaf 0.4856      
CEOL Castro noon/leaf 0.5566 0.0129 0.0101 99.01
  noon/stem 0.5263      
  pm/leaf 0.5970 0.0142 0.0095 105.26
  pm/stem 0.5833      
ARGL Castro noon/leaf 0.5570 0.0248 0.0198 50.51
  noon/stem 0.5610      
  pm/leaf 0.5630 0.0242 0.0187 53.48
  pm/stem 0.6175      
Spectro leaf 0.5505 0.0231 0.0190 52.63
CEME Encino am/leaf 0.5867 0.0184 0.0130 76.92
  am/stem 0.5403      
  noon/leaf 0.5817 0.0184 0.0133 75.19
  noon/stem 0.5666      
  pm/leaf 0.5822 0.0186 0.0134 74.63
Spectro leaf 0.5107 0.0157 0.0150 66.67
GRASS Encino am/leaf+stem 0.1778      

Table 5. Leaf water status: Water Content (WC), Equivalent Water Thickness (EWT expressed in g.cm-2), Leaf Specific Weight (LSW expressed in g.cm-2), and Specific Leaf Area (SLA expressed in cm2.g-1).

CONCLUSION

The premature ending of the first set of passes on the first day was due to the lack of a second battery to power the laptop which drove the spectrometer. The first day only produced 3 passes of the seven flags. We were able to retain a longer battery life by optimizing some procedures, for instance, the laptop was not turned on until the last possible second and was turned off shortly after the last scan of a pass. Efficiency in using the spectrometer quickened the process of acquiring data and using a reverse video mode reduced the energy waste. This allowed us to collect 5 passes at Castro and 6 passes at Encino. We collected more data at Zuma the last day making a total of about 61/2 passes for both days.

The water content measurements were performed at the hotel since there was no battery or power adaptor for the scale. For the next set of measurements, we will use either a portable battery or a car power adapter to power the scale. This will allow us to measure more
leaves with less water loss.