Advances in Modeling Agricultural Systems by Panos M. Pardalos & Petraq J. Papajorgji

Author:Panos M. Pardalos & Petraq J. Papajorgji
Language: eng
Format: epub
Publisher: Springer US, Boston, MA

* * *

Fig. 2 Multispectral image of the two ˜80-ha cotton fields during July 1998. (Image courtesy ITD Spectral Vision, Stennis Space Center, MS.)

* * *

Figure 1depicts that part of a field used during the 1998 variable-rate nitrogen study but here shows the status of the crop during the month of July 1997 [63]. From this aerial view, it is evident that the field is not homogeneous, so the within-field variability we observed on the ground was not surprising. Note the sharp contrast between the mostly green- and yellow-hued areas of the field near the middle of the image and closest to the left edge. In 1996, peanuts were grown in the portion (nearly 43 ha) to the left in the image and appear to have promoted stronger plant vigor (the more green hues) among those cotton plants the following season.

Figure 2 also has the same field as captured in 1998 but now includes the adjacent field of cotton (to the west) that was planted in peanuts during 1997. Notice that although both fields are spatially close and similar in agronomic histories for the year 1998, the spatial pattern of crop vigor (despite differences in color ramping) is different. As was true for 1997, an effect in the following year due to a crop rotation cycle between cotton and peanuts was discernible (more lush) for the west field. Color assignments between 1997 and 1998 were arbitrary and differ due to different choices and selections of color, saturation, and hue in the image processing software [45] and different analysts’ preferences [49].

Both fields (Fig. 2) were planted in cotton May 8, 1998, and then uniformly managed except for four transects in the east field where a variable-rate nitrogen application experiment was conducted. Here, we use for the purpose of illustration only the control and the variable-rate nitrogen treatment based on simulation runs by the cotton growth model, GOSSYM [2]. The simulation model was the integrating vehicle for combining diverse technologies into a precision agriculture application. Each of the 72 plots in this example were about 65 m long, where 18 plots, 8 rows wide, were apportioned to each transect that spanned the length of the field. The crop was harvested during 5–9 October 1998. A four-row CASE-IH cotton picker harvested four rows of each plot. At the end of each plot, the harvester was stopped, the seed cotton dumped into a boll buggy equipped with a scale monitor, and the yield of the plot was weighed and recorded. Additionally, during the crop season, weekly observations were performed on subplots of 5 m in length (5 per transect) embedded within 20 of the 72 GOSSYM managed plots. (This study was accomplished prior to the development of a cotton yield monitor, so other innovations had to be improvised by the research team to obtain the plot yields at harvest.) Another variable-rate nitrogen study in the same vicinity and year, but within other strips of the four transects, provides additional details [60].

Several plant attributes (Table

Download

Copyright Disclaimer:
This site does not store any files on its server. We only index and link to content provided by other sites. Please contact the content providers to delete copyright contents if any and email us, we'll remove relevant links or contents immediately.