Agricultural Management Systems (AMS) and its Grant cooperators conducted tests in the Imperial Valley of California in three production fields of lettuce, flying imagery and taking ground truth frequently, from pre-emergence through harvest.

The primary technology-related goal of this study was to obtain independent verification of the unique performance of the AMS remote sensing system; i.e., to prove that it: (1) detected all stresses (above a small threshold) each time it flew over a field; and (2) did not image "false positives" -- i.e., features in the imagery that appear to depict agricultural stress that are, however, placed into the imagery by "atmospheric clutter" (and are, therefore, of nonagricultural origin). This primary Project goal was achieved.

As far as could be determined, AMS imagery never missed a field problem or condition that stressed the crop. This included stress from fertility problems, plant nutrition problems, insects, diseases, moisture stress, etc. It also showed changes associated with each cultivation practice, included those associated with chemicals applied to the field as well as application and distribution of irrigation water.

Pre-visible stress was detected and confirmed. Analyses of gridded soil samples, taken in a field, corresponded 100% to the stress-patterns shown by AMS imagery.

Finally, as far as could be determined, all stresses depicted by AMS imagery were real stresses, as confirmed by ground truth; i.e., none of them were "false positives".

Unlike a previous study in South Texas, scouting a field with AMS imagery took about the same length of time in this study as normal pattern scouting without imagery. There were two hypotheses advanced to explain this discrepancy: (1) in this study the time required to inspect and interpret the imagery was included in the "scouting time of the field with imagery"; and, (2) because of lack of previous training, scouts were unfamiliar, and therefore slow, in interpreting the imagery and then using it for "imagery-assisted scouting". (As familiarity increased with practice, "scouting time with imagery" decreased.)

It has been noted that AMS imagery did detect bottom rot, but not in its earliest stage. (Bottom rot is a fungal disease that affects the roots and underside-o-leaves of lettuce plants.) AMS has two explanations for this observation: (1) the onset of bottom rot, when it occurs, tends to be very rapid; and, (2) because of high winds, AMS was not able to fly imagery for four days prior to its detection of the bottom rot.

It has been acknowledged, however, that AMS imagery did image the field conditions that led to bottom rot several weeks prior to its discovery, and (even though this was not part of the protocol) the grower was warned at the time of the possibility of bottom rot developing.

Electrical resistivity (i.e., soil salt content) was measured on the ground for AMS by the Imperial Valley Irrigation District using an induction device that they had developed and constructed. There appears to be a high correlation, on bare soils, between electrical resistivity measured by this device on the ground, and AMS imagery taken on bare soils of the same field, or on soils of the field when a low percentage of canopy was present.

Another potentially important observation of commercial applications appears to be a correlation between AMS imagery and soil nitrogen content.

Data for both the resistivity and nitrogen correlations is being analyzed by Mr. George K. Oister and will be provided later as a no-cost addendum.

ACKNOWLEDGEMENTS

We wish to thank the Imperial Valley Irrigation District for contributing ground based soil resistivity measurements of our Project test fields.

We also wish to thank Western Farm Series for providing the Project with free soil analyses on selected fields.