Laserfiche WebLink
MEMO 89.2 <br />The following is an example of a custom classification rule applied to a homogenous parcel, populated <br />mostly by pixels classified as corn: <br />Parcel 1 ~~1 <br />If <br />Crop t'rF°o 1 = `Corrti' <br />.end <br />Crop t~rae 2 = C~r;~ Beans <br />.end <br />Crop t~ ~e 3 = ~J,',~° <br />.~4nd <br />h.~aori~y~ fraction = '~~' of Cann <br />.+"411[1 <br />Homo~~ne~t~f ='~ f~orlacrl~€~ua and ccilti~uous oarcei:~ <br />Then <br />P a rcol 10 01 ~ _ `!:~orrr~ <br />Furthermore, the rules allowed for the detection of parcels that were difficult to identify during the visual <br />interpretation process (i.e., parcels with two or more major crop types with similar temporal NDVI <br />signatures, such as Corn and Sugar Beets). The following example shows the case of a parcel populated <br />by pixels of Corn and Sugar Beets in two equally distributed and homogenous halves: <br />Parcel 1 L~C~0 <br />If <br />Crop t'l~e 1 = 'horn' <br />.And <br />Crap t•yr~e ~ _ ~~u~ar ~~~ts~ <br />.+~nd <br />.~4nd <br />Majoriy~ fraction = 'S25~' of morn <br />.end <br />Horn,o~~onelk~y = ~ ;h•omogenerue and canµgu~ous p~arcpl',~ <br />Then <br />Parcel 10~2C = `~l Parcol' <br />Consequently, the split parcels were flagged and later were manually separated into separate parcels as <br />part of the classification refinement process. <br />In the fourth step, information collected on `Vegetables' and other cover types, such as `Sod Farms' and <br />`Orchard and Tree Nurseries' was manually assigned to the corresponding irrigated parcels. In a similar <br />way, information on irrigation status and crop type collected from interviews with water users, irrigation <br />companies, and irrigation authorities conducted under SPDSS Task 91.3 was incorporated into the <br />classification. During these interviews, water users verified parcel irrigation and crop types on <br />classification maps provided by RTi. For more details on interviews with water users, please see SPDSS <br />Memorandum 91. If water users found discrepancies, the corresponding observations were made on the <br />maps. These observations were then verified against NDVI values and crop type temporal patterns to <br />screen possible errors. After the NDVI verification, these edits were incorporated into the SPDSS crop <br />type classification. Results of the irrigated acreage by crop type are presented in SPDSS Memorandum <br />91. <br />Page 20 of 45 ~R~versfde FecAnotogy, fnc. <br />4'JaYer Resources Errgi~ecr:np an~i CansuFlrnp <br />