RANGE SITE REFLECTANCE CHARACTERISTICS MEASURED BY LANDSAT MSS AND COLOR INFRARED LOW AERIAL PHOTOGRAPHY IN SOUTH DAKOTA MIXED PRAIRIE (REMOTE SENSING, NORTHERN GREAT PLAINS)

A 1982-1984 study using the least square linear prediction model was made to determine the relationships between range biomass (the response variable) and digital reflectance information (the regressor variables) from Landsat MSS and CIR photography. The study sites were located in the continental climate of the mixed prairie of the northern Great Plains (U.S.A.). Elevations ranged from 702 to 778 meters above sea level. Temperatures varied from -34(DEGREES)C to over 38(DEGREES)C. Long term annual precipitation averaged 39.73 cm (1982-1984 averaged 46.26 cm).; Ground truth collections (biomass of total, live and dead components) were made during (1) peak cool-season grass growth (June), (2) peak warm-season grass growth (August), and (3) early winter dormancy (October). Biomass samples were scheduled within ten days of Landsat and aerial photography overpasses.; Selection of regression equations involved (1) F-ratio significance at 0.05 probability level or higher and (2) a coefficient of determination, R('2), equal to or greater than 50%. The test criteria were best satisfied by the June data. The response variable was the live biomass (BLVSQ and BL) components. The regressors with the higher R('2)s were the near infrared (LS6, LS7) and red band (LS5) ratios from Landsat MSS. The prediction fit models to the data were: (1) BLVSQ = -56668 + 44697 LS6/LS5 (R('2) = 61.0%), (2) BLVSQ = -43856 + 37042 LS7/LS5 (R('2) = 62.0%), (3) BL/BD = -12.9 + 7.95 LS6/LS5 (R('2) = 53.5%), and, (4) BL/BD = -10.0 + 6.30 LS7/LS5 (R('2) = 49.6%).; The square root of the solution to either equation (1) or (2) produces the live biomass (BLV) component (g/m('2)) related to the June sampling period in western South Dakota. Substitution of that component (BLV) into either equation (3) or (4) yields the dead biomass component. The sum of live plus dead components produces total biomass (g/m('2)). Total biomass multiplied by 10.00 converts into kg/ha, a managerial unit.