Identification of irrigation practices using photographic and optical-mechanical scanning remote sensing techniques

Graduation date: 1981Irrigation is essential for profitable agriculture in the western\ud United States. It is the largest consumer of water and power in\ud Oregon. Conflicting uses of water and power and their scarcity demands\ud judicious planning for allocation of these resources. Creditable\ud baseline data are not only needed for irrigated crop acreages, but\ud also for irrigation methods as an indicator of amount of water and\ud power used. A data-gathering system is needed which is quick, reliable\ud and comparatively inexpensive.\ud The primary objective of this research was to develop a\ud methodology to use remotely sensed data for irrigation method\ud identification. The main tasks were to: (1) analyze multidate images\ud for different emulsions and scales to identify irrigation methods:\ud (2) develop mission parameters; (3) develop an image interpretation\ud key; (4) investigate the feasibility of using a quantitative approach\ud to irrigation method identification.\ud The North Unit Deschutes Irrigation District of central Oregon\ud was selected to develop and test the methodology. Field layout of\ud different irrigation methods, their association with crops, soils, and\ud slopes were noted for the area imaged by three CIR aerial flights\ud during the 1979 growing season. A detailed crop calendar showing\ud different phenological stages that affect the spectral properties of\ud crops was developed.\ud Images used for analysis included: Landsat MSS color transparencies\ud at 1:1,000,000 for June 4, June 22, July 18, 1979; Landsat MSS\ud color prints at 1:250,000 for May 17, and June 22, 1979; Landsat RBV\ud images at 1:125,000 for June 8, 1978 and July 28, 1979; U-2 CIR\ud photographs at 1:30,000 for June 28, 1973 and at 1:130,000 for August\ud 2, 1978; CIR aerial photographs for May 12, July 9, and August 5,\ud 1979 at 1:30,000 and 1:23,000.\ud Two methods of analysis were used. Manual interpretation of the\ud images employed light tables, magnifiers and stereomicroscope. Each\ud test field along the flight line was carefully studied for detection\ud of irrigation patterns using associated tones, colors, textures, and\ud their temporal variations, in addition to crop calendar data, soil,\ud and slope maps, ground information gathered, and black and white\ud images at 1:8,000. Digital analysis of the data for discrimination\ud of irrigated and non-irrigated crop types was discouraging as the data\ud were for the late growing season (July 28, 1979), when most crops were\ud either harvested or in the senescing stage and their spectral\ud signatures overlapped.\ud Conclusions drawn from this study are that: image-oriented\ud analysis rather than quantitative analysis is best suited for irrigation\ud method discrimination; CIR images provide better discrimination\ud than black and white images; while irrigated and non-irrigated crops\ud can be discriminated using multidate 1:1,000,000 images, for\ud satisfactory discrimination of all irrigation methods, 1:30,000 or\ud larger imagery is needed; and the best time for differentiating\ud irrigation methods is the early growing season.\ud As a result of this research, optimum mission parameters for\ud flying photographic missions were developed and an image interpretation\ud key useful for irrigation methods discrimination was developed