View angle effects on relationships between MISR vegetation indices and leaf area index in a recently burned ponderosa pine forest
While nadir-viewing passive multispectral sensors have limited utility for characterizing the full dimensionality of forest canopies, multi-angle remote sensors such as the Multi-angle Imaging SpectroRadiometer (MISR) may improve detection of canopy architecture, canopy cover, and leaf area index (L...
Gespeichert in:
Veröffentlicht in: | Remote sensing of environment 2007-03, Vol.107 (1), p.322-333 |
---|---|
Hauptverfasser: | , , , |
Format: | Artikel |
Sprache: | eng |
Schlagworte: | |
Online-Zugang: | Volltext |
Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
Zusammenfassung: | While nadir-viewing passive multispectral sensors have limited utility for characterizing the full dimensionality of forest canopies, multi-angle remote sensors such as the Multi-angle Imaging SpectroRadiometer (MISR) may improve detection of canopy architecture, canopy cover, and leaf area index (LAI) of forest canopy versus understory vegetation. Our objective was to determine whether data from the MISR sensor could improve estimates of LAI across a post-fire ponderosa pine forest located in the Black Hills of South Dakota. We measured LAI during the 2002 and 2003 growing seasons and created continuous LAI maps using Landsat TM and ETM+ data (mean
R
2
=
0.81). We fit linear regression models of total and canopy LAI, using a series of MISR at-nadir and off-nadir or anisotropic vegetation indices as predictor variables, for each of five sampling periods. We found the best LAI model fits using either a new hotspot-adjusted normalized difference vegetation index, NDVI
HS, NDVI calculated from the −
60° view angle, or NDVI at-nadir and either the Hotspot–DarkSpot Index (HDS) or the normalized difference anisotrophic index (NDAX) (
R
2
=
0.56–0.91). However, differences in fits of these best models and those including NDVI at-nadir ranged from only 1 to 8%. Reflectance anisotropy patterns related strongly to understory vegetation phenology. We found that the relationships among NDVI or the enhanced vegetation index (EVI) and canopy or total LAI showed little variation across view angles when the understory vegetation was senesced and significant anisotropy when understory green LAI was greatest. These findings demonstrate the value of multi-temporal measurements during periods of understory phenological change, even if the overstory LAI is relatively stable. We also evaluated the performance of the MISR LAI product and found moderate fits between the MISR LAI and our field- and Landsat-derived canopy LAI (
R
2
=
0.21–0.44). |
---|---|
ISSN: | 0034-4257 1879-0704 |
DOI: | 10.1016/j.rse.2006.06.019 |