There are several methods which can be used to measure a crop's
acreage in a scene. Techniques and procedures which have been
developed and tested in LACIE operations or in its research and devel-
opment programs have been used by USDA analysts to estimate crop
acreages. An effort was made by the analysts to develop procedural
and processing alternatives relative to different wheat growing
regions and variations in cultural/cropping practices in several
countries. Each of the examples presented support a hypothesis that
analyst perceived conditions by region and year dictate the multi-
spectral analysis decision logic.
II. Conditions Encountered in Landsat Analysis
User goals also dictate to a great extent what type of operational
environment must be established in order to perform the required
analysis. Remote Sensing analysis demands that there exists an
integrated procedure for utilizing personnel and equipment in an
efficient manner. Equipment and personnel have to compliment each
other in order to establish a set of processing procedures which
are needed to perform the required spectral data analysis. It is
advantageous to use electronic data processing equipment to analyze
conditions which are fixed in a scene and are repeated from year
to year. Conversely, the variable conditions and characteristics
are more conducive to analyst interpretation. It is therefore
necessary to group scene characteristics into those which are vari-
able and those which are fixed. Those scene characteristics which
are fixed should be used as a guideline in developing procedures to
handle the various fixed attributes. In addition to a perceived
methodology for exploiting Landsat data, its practical utility
depends upon agricultural considerations, available processing tech-
niques and finally the optimum merging of considerations and techniques.
A. Agricultural Considerations
Each year farmers decisions both fixed and variable are represented
in remotely sensed Landsat data. Within agriculture there are many
conditions which exist that force farmers to choose between various
alternatives in order to maximize their return. One of the most
dominant factors farmers have to consider are the existing climatic
conditions. This factor strongly effects which crops a farmer can,
or will choose to plant. However, the farmers also have to take into
consideration factors such as soil type, topographic conditons, plant
varieties, property lines, fertilization practices, and economic
conditions. All such factors give the farmer choices in three areas:
what to plant, how much to plant and how to care for what is planted.
It is primarily the results of the first two choices, what and how
much to plant, which can be seen on the Landsat images. The various
plant types can be distinguished by nominal planting and harvesting
dates, and identifying the spectral history of the crop through the
season. The problem of identifying crop types is closely related to
— TT CD
uU oT m -—
TT eh eh E RE DVD UN
Z3 € QD £f S -3 (NH — *€ 4 CT och fU U) (D +
~~ etre SM OCS