Relating field observations of plant phenological events to remotely sensed depictions of land surface phenology remains a challenge to the vertical integration of data from disparate sources. We capitalized on legacy datasets pertaining to reproductive phenology and biomass in conjunction with analysis of moderate resolution Landsat Thematic Mapper (TM) imagery at the Jornada Basin (JRN) Long-Term Ecological Research site in the northern Chihuahuan desert in New Mexico. We set out to answer the following questions: (1) How can field measurements of reproductive phenology for dominant grass and shrub species enhance depictions of land surface phenology using Landsat 5 TM imagery?; and, (2) Does Landsat 5 TM imagery serve as a reliable and accurate proxy for vegetation biomass in this highly heterogeneous arid ecosystem? Field data from fifteen long-term research sites representing five distinct vegetation communities include aboveground plant biomass measured three times annually and counts by phenophase measured monthly. Twenty-one TM images were corrected for atmospheric effects to obtain surface reflectance for deriving soil-adjusted vegetation index (SAVI) values depicting vegetation greenness. This pilot study involves evaluating patterns in biomass, phenology, and vegetation greenness from May 2006 to Nov 2009 encompassing two anomalously wet years (2006 and 2008).

Results/Conclusions

Strong seasonal patterns in vegetation greenness from TM reflected dynamics in aboveground plant biomass. TM-based responses in fall 2008 production to 2006 and 2008 above-average rainfall were consistent across all sites. The cumulative effect of 2006 and 2008 rainfall was most prominently reflected in biomass at mesquite-dominated sites. Biomass ranged widely over time and across sites (0.1 g m-2 to 422.2 g m-2).  SAVI effectively tracked increases in biomass across all plant communities although values exhibited a low dynamic range across large differences in plant biomass. Field data regarding reproductive phenology constitute counts for given phenophases by species. Such measures of abundance are most informative to remotely sensed interpretations of phenology where there are clear dominant species. Nonetheless, the field perspective is historically unrivaled for highlighting species responses to climate drivers and is being evaluated further. The range in vegetation physiognomy and productivity at the JRN LTER is well-suited for field validation exercises for land surface phenology and retrieval of biophysical parameters via remote sensing. Plant responses to changes in climate (e.g., temperature and amount/seasonality of rainfall) manifest through phenological patterns. Therefore, field-tested protocols for monitoring plant phenology hold great promise for landscape monitoring and quantifying ecosystem responses to climate change.