TITLE:

Use of an aerial Hyperspectral Sensor to Diagnose Stress and Predict Yield at AG-CARES, Lamesa, TX, 2001.

AUTHORS:

Terry Wheeler (TAES), Harold Kaufman (TCE), Wayne Keeling (TAES), and Randy Boman (TCE), Associate Professor, Extension Specialist, Professor, Extension Specialist..

RESULTS AND DISCUSSION:

A hyperspectral sensor was purchased by the Lubbock and Amarillo Research and Texas Cooperative Extension Centers during 2001. The AG-CARES circle was flown using the sensor on 7 September. While the unit does have global positioning system capabilities, it was not functional during this flight. This means that small areas or points could not be accurately identified from the images. The image was therefore analyzed with respect to large plots or easily recognized areas. The measurements investigated included fertility (phosphorous in the long-term minimum tillage-cotton area); variety test (conventional tillage long term-cotton); and yield as measured by a monitor in areas 1A (conventional tillage) and 1B (part of a 3-year cotton/narrow cotton/peanut study). Only the variety test showed a relationship reflectance as seen here by the different shaded plots in the hyperspectral image (Fig. 1). Fiber Max and Stoneville varieties had a different pattern of reflectance than did Paymaster 2326BGRR, PM 2326RR, or PM 2266RR. The longer season varieties of Fiber Max and Stoneville (depicted by FM 989BGRR) had lower reflectance over the blue, green, and red spectrum (431 - 700 nm), but a higher reflectance over the near-infrared spectrum (710 - 875 nm) than did PM 2326RR (Fig. 2).

Figure 1.

Figure 2.