Laserfiche WebLink
MEMO 93 <br />Irrigated Lands <br />Classification (Mask) <br />Multitemporal Landsat <br />Image Composite (All Bands) <br /> <br />Masked Multitemporal <br />Landsat Image Composite <br />Step 3: <br />Post Classification <br />Refinement <br />Signature Extraction and <br />Analysis <br />Preliminary Irrigated <br />Per Pixel Crop Type Classification <br />Step 4: <br />Classification <br />Review <br /> <br /> <br />Cro s GIS <br />Acreage <br />1957 GIS <br />Acreage <br />1976 <br />Alfalfa 224,792 229,545 <br />Corn 399,207 478,710 <br />Beans 89,027 65,606 <br />Grass 246,126 353,833 <br />S.Grains 119,605 47,467 <br />S. Beets 48,305 80,159 <br />Totals 1,127,061 1,255,320 <br />QA/QC and <br />Classification Review <br />Step 1: <br />Irrigated <br />Lands <br />Masking <br />Step 2: <br />Spectral Signature <br />Analysis & <br />Supervised <br />Maximum <br />Likelihood <br />Pixel Classification <br />Delineated parcel <br />boundaries <br />Figure 7. Historical irrigated crops classification procedure. <br />The per image mean NDVI and standard deviation value for each of the 50 classes were then exported to <br />Microsoft Excel and graphed. Mean NDVI and standard deviation spectral signatures from 2001 FSA <br />reference data were added to the graph for comparison. Figure 8 is an example of this graphical <br />comparison, and illustrates how 2001 FSA reference data helped to identify, in this instance, four of the <br />Page 13 of 59 ,Rlreradde ~eshn~P~gy, ~os~ <br />Irrigated Crop <br />Classification by parcel <br />