Landsat Soybean Field Yield Prediction / Phosphorus and Potassium Removal Mapping
Related pages: Landsat Correlation to Soybean Yield Interpolated Landsat Correlation to Yield
Images transition every 4 seconds (or click arrows or numbers); description appears below graphic. (Only pixels that completely represent surface within the field boundary should be used.)
Soybean yield zones in Slide 4 above can be converted to phosphorus removal based on a "newly revised" removal rate of .61 pounds per bushel (SDSU, 2014) to produce zones, from higher to lower, of 32.9, 30.3, and 27.6 pounds per acre.
Soybean yield zones in Slide 4 above can be converted to potassium removal based on a removal rate of 1.2 pounds per bushel (Mallarino and Sawyer, 2013) to produce zones, from higher to lower, of 64.7, 59.5, and 54.3 pounds per acre.
(Removal rate can be based on any requested conversion source. Recent and lowest values found were applied here).
If applied correctly, Landsat can correlate highly to spatial patterns of soybean yield. GIS Ag Maps can additionally calibrate Landsat values to corn yield to produce phosphorus and potassium removal maps; maps can be classified into zones (based on natural grouping of the data, or continuous. Contact GIS AgMaps with question or more information.
This page shows a method for coherent, Landsat-based interpolated corn yield prediction mapping based on valid pixels. Landsat NIR reflectance of soybean fields during mid R-stages has been shown to highly correlate to spatial patterns of yield (Hollinger, 2011); NIR solely correlates to soybean yield higher than indices such as NDVI, GNDVI, and others. For soybeans, visible bands should not be used for yield prediction during R-stages because they saturate; including visible band data, such as the red band in NDVI or green band in GNDVI, will lower correlation with yield as is shown Correlation to Soybean Yield page. Landsat can be used to develop a corn yield prediction map where only valid pixels are used and extended to the field boundary; the map can be in continuous form as shown in the slideshow or can be classified into zone as shown shown after. The final map can be calibrated (converted) to yield amounts based on an equation or any specified yield range and/or average value (described below); this way you can have an estimation of how much higher and lower yield is in different areas of the field.
Steps that show how Landsat can be used to produce a yield prediction map of continuous data or zones are included below (same process for any crop; soybean field shown below) and can be seen in different locations of the website including the homepage slideshow. The map can be produced to a field extent of different shapes and sizes and can be calibrated with yield values based on an equation or a specified yield range and/or average value; the data is then essentially a generalized yield map.
The progression from raw Landsat imagery to maps to a field extent (or any other extent) is the same for any crop. From left to right the steps are: 1) produce Landsat with pixels that represent correct value to predict yield well enough; 2) use only pixels that represent the crop, not pixels that average in surface outside the field extent or major non-crop surface within the field perimeter; 3) data is interpolated or resampled to a finer resolution for a more coherent map (below, the pixels have been modified from the 30 x 30 meter native resolution to a one-meter resolution); 4) zones can be developed (the appropriate classification method, amount, and minimum size of zone need to be determined).
Calibration of Landsat Maps to Yield
Landsat yield prediction maps can be calibrated to yield based on solely a field average or a specified yield range and average can be applied. In either case, the calibrated map will keep the same proportions to the Landsat value map (in other words, the map symbology will look the same before and after calibrated to yield).
If just a field average is known, equations from Hollinger (2011) can be applied that predict yield for the entire field. NIR reflectance values are necessary to apply the equation. The equation predicts normalized (to the mean) yield amounts based on normalized (to the mean) NIR reflectance. Once normalized yield values are calculated they are multiplied by a field average to acquire a yield prediction map. To produce the equation for NIR, pixel groups from many fields over four seasons from mid R-stages image dates were normalized to the mean and combined into one plot where they were correlated with corresponding clean yield monitor data (R² = 0.63; n = 3,807 pixels). The yield variability prediction model is based on fairly typical seasons.
Maps can also be calibrated based on a specified yield range and field average.
Reference
Hollinger, D. 2011. Crop Condition and yield prediction at the field scale with geospatial and artificial neural network applications. Dissertation. Kent State University.
Mallarino, A.P. and J.E. Sawyer. 2013. Update to Iowa phosphorus, potassium and lime reccomendations. Integrated Crop Management News. Iowa State University; Extension and Outreach. Published originally on 9/21/2013. Cited at: http://www.extension.iastate.edu/CropNews/2013/0920mallarinosawyer.htm
SDSU. 2014. iGROW. Recent news: Building Soil Phosphorus? Posted March 6th, 2014. South Dakota State University, Brookings, SD 57007. Cited at: http://igrow.org/news/building-soil-phosphorus/#sthash.wZTNH9vW.dpuf