Data Package Metadata   View Summary

The PLS Surveys were conducted from 1816-1907 in Michigan. As described in Goring SJ, Mladenoff DJ, Cogbill CV, Record S, Paciorek CJ, Jackson ST, et al. (2016) Novel and Lost Forests in the Upper Midwestern United States, from New Estimates of Settlement-Era Composition, Stem Density, and Biomass. PLoS ONE 11(12): e0151935. doi:10.1371/journal.pone.0151935, "Michigan surveyor observations were transcribed to Mylar sheets overlaid on State Quadrangle maps, so that the points were displayed geographically, and then digitized to a point based shapefile (Ed Schools, pers.comm.; Great Lakes Ecological Assessment. USDA Forest Service Northern Research Station. Rhinelander, WI. http://www.ncrs.fs.fed.us/gla/)". Students, with the help of Simon Goring at the University of Wisconsin, Madison in the Williams lab transcribed the data from the Mylar maps to a GIS shapefile. That version 0.1 GIS shapefile was sent to Jody Peters at the University of Notre Dame where she worked with Charlie Cogbill to conduct a number of quality control checks which updated the data for southern Michigan.

Data Processing Steps going from the original file sent from UW through the final processing at Notre Dame (the final version number from the processing was 1.6, but the file was renamed as version 1.0 for the official archiving of the data pacakge msb-paleon.30) 1.Simon sent the so_michigan.shp which had 33,719 records (or corners). We updated the file to provide information for each corner on whether 1) corners were exterior on a township border or on the interior of the township, 2) labeled corners as ¼ section or sections corners, and 3) what county each corner was in. 2.Removed corners that were in the northern Michigan database that were already a part of the Upper Midwest Files. 3.Removed corners with identical tree data found at the same location (used the uniqueID attribute column to find corners with identical tree data). 4. Removed corners that were on the west border of townships 5.Charlie went through the data and made corrections when there were obvious errors such as when the data was shifted over a column. He also calcuated new azimuths for each tree (AZ1_360, AZ2_360, AZ3_360, AZ4_360) that incorporated information about the quadrant the tree was located in. So if the quadrant number was 1, the AZ as read from the mylar maps was used as is. If the quadrant number was 2 Charlie took 180-the WI AZ. Quadrants of 3 were 180+the WI AZ and quadrants of 4 were 360-the WI AZ. In the Azimuths Charlie also changed the “9999” standing for missing values to be “0”. NOTE from Charlie about the general QA/QC that has been done for the corners to the East of the Meridian in southern MI (that is any entry with with an "E" in the range column): First, if azimuth is missing -- 9999 or comment "missing", but not 0-- and the quadrant is known, a 45 bearing is inserted which preserves the quadrant and "averages" the bearing. If the bearing is >90, --usually double hits or overlapping fields-- the obvious typographical errors were corrected otherwise the corner is discarded. Second, trees >100" had numerals redistributed over clearly empty fields or the corner discarded. Third, all 0 bearings were scanned for notes in the comment field to assign cardinal directions, that is true 0 to 360 azimuths. Fourth --most importantly due to numerous large diameters up to 89" in Detroit region-- any tree with a remaining 0 bearing was assumed to have corrupted diameter and azimuth fields and the corner discarded. This is very conservative as bearing errors do not affect density, and diameter errors were eliminated for the least reason. 6. The final version of the data have 29,278 corners

Additional details can be found in the methods sections in: Paciorek et al., 2019, ADD TITLE AND BIOARCHIVE LINK. Code for the data processing and analyses can be found on GitHub: https://github.com/PalEON-Project/PLS_products.