Navigation path


Path: RAMI-IV : EXPERIMENTS : ACTUAL CANOPIES : WELLINGTON CITRUS ORCHARD : Thresholded hemispherical photograph
EXPERIMENTSRESULTSMODELSPARTICIPANTS Up Print version Decrease text Increase text Home

Thresholded hemispherical photograph (binary fisheye image): thp

This measurements involves determining the binary probability (i.e., 0 or 1) of seeing the sky when looking up into a particular direction from a given location at the bottom of the canopy. In order to simulate results similar to those obtained from hemispherical photographs the thp measurement requires to simulate 250000 directions uniformly distributed throughout the upper hemisphere. If a given RT model is not able to simulate individual directions then solid angles should be chosen that are as small as possible (and this value should then be reported to the RAMI coordinators via a report files). More specifically,

  • For RT models that generate deterministic canopy representations on the basis of finite-sized objects (geometric primitives) the output of this measurement is thus 1 for all directions that enable to see the sky unhindered and 0 for all those directions where one or more objects prevent an unobstructed view of the sky.
  • For RT models that use statistical representations of crown volumes (i.e., pseudo-turbid medium) the output of this measurement will be 1 for all directions that allow to see the sky unhindered and 0 for all those directions where the sky is obstructed by an object. However, since the probability of transmission through a pseudo-turbid medium may fall anywhere in the range from zero to unity a threshold will have to be applied to determine when the sky is obstructed in a given view direction. Although RAMI participants are free to decide on the actual value of this threshold, the most obvious choice is to declare all directions for which the transmission probabilities are smaller than unity as being obscured/blocked.

The thp measurement applies to all ACTUAL canopy scenes and is to be carried out at nine different locations within these RAMI scenes (labelled A to I). At each of these locations the height above the background at which the measurements is to be carried out is 0.05 m. Nevertheless, the exact position of the thp measurement locations will depend on the whereabouts of the origin of the coordinate system in the scene:

  1. RAMI scenes that have the origin of the coordinate system in the lower left corner of the scene: feature the following x,y,z locations for the thp measurements:
    • location A = 20.0,20.0,0.05
    • location B = 20.0,50.0,0.05
    • location C = 20.0,80.0,0.05
    • location D = 50.0,20.0,0.05
    • location E = 50.0,50.0,0.05
    • location F = 50.0,80.0,0.05
    • location G = 80.0,20.0,0.05
    • location H = 80.0,50.0,0.05
    • location I = 80.0,80.0,0.05
  2. RAMI scenes that have the origin of the coordinate system in the center of the scene: feature the following x,y,z locations for the thp measurements:
    • location A = -30.0,-30.0,0.05
    • location B = -30.0,0.0,0.05
    • location C = -30.0,30.0,0.05
    • location D = 0.0,-30.0,0.05
    • location E = 0.0,0.0,0.05
    • location F = 0.0,30.0,0.05
    • location G = 30.0,-30.0,0.05
    • location H = 30.0,0.0,0.05
    • location I = 30.0,30.0,0.05

At each of the above nine locations in a given RAMI scene, the thp measurements are to be taken at a height of 0.05 m and for 250000 directions that are uniformly distributed across the upper hemisphere. This number is but a fraction of the pixels that are available in actual digital cameras currently used for field inventories! An ASCII file with the exact (zenith angle & azimuth angle) directions can be found here. In order to be able to carry out the thp measurements at all angles in the upper hemisphere the finite-sized RAMI scenes are to be added an infinite number of times to the outsides of themselves (in other words cyclic boundary conditions apply).

thp file-naming convention:

Since thp measurements are independent of wavelength and in order to specify the location (A - I) of the measurement within the canopy scene the RAMI file-naming convention will have to be adapted. Therefore, please assign, B00 as the band identifier, and replace the solar zenith angle identifier by a zero followed by the letter indicating the thp location (e.g., 0A) and - where applicable use as solar azimuth angle tag simply 000 in the measurement filename. This naming strategies then results in the following filenames, for example, thp_raytran-HET07_JPS_SUM_B00_0C.mes for a thp measurement at location C in the Jaervselja pine stand and thp_raytran-HET14_WCO_000_B00_0H.mes for a thp measurement at location H in the Wellington citrus orchard.

Within RAMI-IV the thp measurement applies only to a selection of canopy scenes:

    Formatting of measurement-file content:

    Header line content # rows # columns in file hemisphere tag
    Header line format integer integer integer
    Body lines content VZA [rad] VAA [rad] transmission probability
    Body lines format long-real long-real integer


    • VZA: Viewing Zenith Angle
    • VAA: Viewing Azimuth Angle
    • Hemisphere tag: Set to 1 to indicate upward looking instrument (0 would indicate downward looking instrument).
    • A long-real is defined as having 10 digits after the decimal point, e.g., 1.0000000000

    Example output (first few lines for a dirgapf measurement in a determinstic canopy scene):

    250000 3 1
    1.5694883253 0.0048036585 0
    1.5694883253 0.0144109755 0
    1.5694883253 0.0240182925 0
    1.5694883253 0.0336256094 1
    1.5694883253 0.0432329264 1
    1.5694883253 0.0528402434 0
    1.5694883253 0.0624475604 0
    1.5694883253 0.0720548774 0
    1.5694883253 0.0816621944 1
    1.5694883253 0.0912695113 0
    1.5694883253 0.1008768283 1
    1.5694883253 0.1104841453 0
    1.5694883253 0.1200914623 0

    Right navigation