(or 
nd 
of 
oil 
[15]. 
at 
e 
976, 
olled 
a 
the 
ctral 
rary 
tive 
ted 
omputer 
ty 
from 
d in 
4 um 
merged 
apes. 
agery 
r 
craËt 
0 
ystem. 
y the 
ere 
rmat. 
  
Landsat multispectral scanner data were previewed using black and 
white transparencies of the image for each band to establish data 
quality and cloud cover conditions within the intensive test sites and 
the complete Landsat frames. Following data quality evaluation, 
Landsat multispectral scanner data were processed to 9-track computer 
compatible tapes. 
Data Library 
The multispectral data library maintained at Purdue/LARS for the 
LACIE Field Measurements Project contains over 100,000 spectra (cor- 
responding to measurements of over 800 plots and fields) and over 
2,000 observations made with Landsat-band radiometers [16]. The 
library also includes several hundred scenes of aircraft and satellite- 
acquired scanner data. A data library catalog was prepared for each 
crop year containing summary and detailed schedules of data acquisition 
by location, sensor system, and mission. Digital data products available 
for analysis include Landsat and airborne scanner data, helicopter- and 
truck-spectrometer spectra and ancillary data, and tripod-radiometer 
spectra and ancillary data. Aerial photography concurrently with 
spectrometer and scanner data are also available. 
Data Analysis Systems 
LARSYS (Version 3.1) is a fully documented software system designed 
to provide the tools for analysis of multispectral scanner data [17]. 
The pattern recognition and interactive data handling techniques in 
LARSYS have been used world-wide for analysis of the aircraft and Landsat 
scanner data for many applications. 
EXOSYS is a specialized software system developed at LARS for 
analysis of spectrometer data. It provides researchers with the 
capability to recall spectrometer data by sorting on combinations 
of measurement (e.g., solar elevation) and ancillary parameters (e.g., 
leaf area index). Analysis features of EXOSYS include the ability to 
compute functions of band-averaged reflectances and perform correlations 
with crop parameters, polynomial curves may be fitted to the data using 
the least squares technique. Initial results are reviewed and then 
sent to a line printer or a graphics plotter. 
Summary 
A comprehensive set of multitemporal spectral, agronomic, and 
meteorological data were acquired for three test sites in Kansas, 
South Dakota, and North Dakota for three years. Spectral measurements” 
were made of controlled, experimental plots of wheat using truck-mounted 
spectrometers and of commercial fields of wheat and other crops by a 
helicopter-borne spectrometer, an airborne scanner, and Landsat MSS. 
The spectral data are calibrated to provide valid comparisons of data 
from different sites, sensors, and dates, and are supported by an