A Project Linking In-situ and Satellite Measurements to Validate MODIS Terrestrial Ecology Products arren B. Cohen, US Forest Service; Stith T. Gower, University of Wisconsi P. Turner, Oregon State University; Peter B. Reich, University of Minnes Steven W. Running, University of Montana
22
Embed
A Project Linking In-situ and Satellite Measurements
A Project Linking In-situ and Satellite Measurements to Validate MODIS Terrestrial Ecology Products. Warren B. Cohen , US Forest Service; Stith T. Gower , University of Wisconsin; David P. Turner , Oregon State University; Peter B. Reich , University of Minnesota; - PowerPoint PPT Presentation
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
A Project Linking In-situ and Satellite Measurementsto Validate MODIS Terrestrial Ecology Products
Warren B. Cohen, US Forest Service; Stith T. Gower, University of Wisconsin;
David P. Turner, Oregon State University; Peter B. Reich, University of Minnesota;
Steven W. Running, University of Montana
Objectives
• Develop better understanding of the climaticand ecological controls on total net primaryproduction and carbon allocation within andamong biomes
• Learn how flux tower-measured NEE andfield-measured NPP co-vary in time & how totranslate between them using ecological models
• Explore errors and information losses thataccrue when extrapolating field data to coarse-grained (1 km) surfaces
• Provide high quality site-specific data layersat four sites that can be compared to MODISand other sensor products
Technical
Scientific
Sites
BOREAS Northern Old Black Spruce (NOBS)
Muskeg (open black spruce),“Closed” black spruce, Aspen,Wetlands, Jack pine
Meeting Our Land Cover Mapping NeedsAGRO• corn (class--label fields in-situ)• soybean (class--label fields in-situ)• other (classes--label clusters in-situ & HD imagery)
HARV• hardwood/conifer (relative proportions--from HD imagery)• other (classes--label clusters in-situ & HD imagery)
KONZ• grass (short/tall combined class--labels from plots & HD imagery)• forest (one class--labels from plots & HD imagery)• shrub (percent--from HD imagery calibrated with camera observations, plots)• other (classes--label clusters in-situ & HD imagery)
HD (high definition) imagery: ADAR, IKONOS, AVIRIS mission photos, MQUALS
NOBS• conifer/hardwood/standing dead (relative proportions--from camera observations, HD imagery)• “ground” cover (relative proportions--from camera observations, moss classes from ocular estimate)• other (classes--label clusters in-situ, camera observations, & HD imagery)
“ground” cover classes: moss, lichen, herbaceous, shrub, fine litter, tree regeneration, coarse wood debris, water
Leaf-off HARV
Capturing seasonalityWith ETM+ is important to both landcover and LAI mapping
Land cover (e.g.,…)• aspatial: compare frequency distribution of translated site-specific classes with same from MODLand• spatially explicit: summaries of site-specific cover proportions within MODLand- labeled cells
LAI/fPAR (e.g.,…)• mean 1 km cell values vs. MODLand values• distributions of fine-grained values within MODLand cells
NPP (e.g.,…)• integrated 1 km cell values vs. MODLand values• distributions of fine-grained values within MODLand cells• spatially degrade land cover and LAI, repeat modeling, redo above NPP comparisons• informationally degrade land cover, repeat modeling at fine grain, redo comparisons