nt ti m
S-5 eley, 130 Mulford Hall, #3114, Berkeley, CA 94720-3114, USA a r t i c l e i n f o
Received 25 October 2013
Received in revised form 29 March 2014
Accepted 2 April 2014
Available online 3 May 2014
Field spectroscopy th rates of net primary intensive agricultural n, 2009). The roots and on and peat formation
Remote Sensing of Environment 149 (2014) 166–180
Contents lists available at ScienceDirect
Remote Sensing o j ourna l homepage: www.edefined as wetland ecosystems with greater than 40 centimeters (cm) of surface organic matter, which represent 16–33% of the global soil carbon pool (Bridgham, Megonigal, Keller, Bliss, & Trettin, 2006). (Moore, 1987; Rasse, Rumpel, & Dignac, 2005), resulting in long-term carbon sequestration, soil stability and accretion (Gorham, Lehman,
Dyke, Clymo, & Janssens, 2012; Miller, Fram, Fujii, & Wheeler, 2008;Wetlands cover just between 2% and 6% of the earth's land surface, but store a large proportion of the world's carbon in terrestrial soil reservoirs (approximately 15 × 102 petagrams) (Kayranli, Scholz,
Mustafa, & Hedmark, 2010). Most of this carbon is found in peatlands, freshwatermarshes is particularly productive, wi productivity as high as that of tropical forests and ecosystems (Miller & Fujii, 2010; Rocha &Goulde rhizomes of this vegetation drive soil organic carb1. Introduction These significant carbon sinks are primarily a result of on-site (autochthonous) plant production (Moore, 1987). Emergent vegetation of⁎ Corresponding author. Tel.: +1 650 329 4279.
E-mail addresses: firstname.lastname@example.org (K.B. Byrd), jessica. (J.L. O'Connell), email@example.com (S. Di Tommaso (M. Kelly). http://dx.doi.org/10.1016/j.rse.2014.04.003 0034-4257/Published by Elsevier Inc.that will identify variability in small, fragmented marshes common to the Sacramento–San Joaquin River Delta and elsewhere in the Western U.S.
Published by Elsevier Inc.Multispectral sensor
Error reportinga b s t r a c t
There is a need to quantify large-scale plant productivity in coastalmarshes to understandmarsh resilience to sea level rise, to help define eligibility for carbon offset credits, and tomonitor impacts from land use, eutrophication and contamination. Remote monitoring of aboveground biomass of emergent wetland vegetation will help address this need. Differences in sensor spatial resolution, bandwidth, temporal frequency and cost constrain the accuracy of biomass maps produced for management applications. In addition the use of vegetation indices to map biomass may not be effective in wetlands due to confounding effects of water inundation on spectral reflectance. To address these challenges, we used partial least squares regression to select optimal spectral features in situ and with satellite reflectance data to develop predictive models of aboveground biomass for common emergent freshwatermarsh species, Typha spp. and Schoenoplectus acutus, at two restoredmarshes in the Sacramento–San Joaquin River Delta, California, USA. We used field spectrometer data to test model errors associated with hyperspectral narrowbands andmultispectral broadbands, the influence of water inundation on prediction accuracy, and the ability to develop species specific models. We used Hyperion data, Digital GlobeWorld View-2 (WV-2) data, and Landsat 7 data to scale up the best statistical models of biomass. Field spectrometer-based models of the full dataset showed that narrowband reflectance data predicted biomass somewhat, though not significantly better than broadband reflectance data [R2 = 0.46 and percent normalized RMSE (%RMSE) = 16% for narrowband models]. However hyperspectral first derivative reflectance spectra best predicted biomass for plots where water levels were less than 15 cm (R2 = 0.69, %RMSE = 12.6%). In species-specific models, error rates differed by species (Typha spp.: %RMSE= 18.5%; S. acutus: %RMSE= 24.9%), likely due to themore vertical structure and deeperwater habitat of S. acutus. The Landsat 7 dataset (7 images) predicted biomass slightly better than the WV-2 dataset (6 images) (R2 = 0.56, %RMSE = 20.9%, compared to R2 = 0.45, RMSE = 21.5%). The
Hyperion dataset (one image) was least successful in predicting biomass (R2 = 0.27, %RMSE = 33.5%). Shortwave infrared bands on 30 m-resolution Hyperion and Landsat 7 sensors aided biomass estimation; however managers need to weigh tradeoffs between cost, additional spectral information, and high spatial resolutionb Department of Environmental Science, Policy and Management, University of California, BerkEvaluation of sensor types and environme biomass of coastal marsh emergent vegeta
Kristin B. Byrd a,⁎, Jessica L. O'Connell b, Stefania Di Tom a Western Geographic Science Center, United States Geological Survey, 345 Middlefield Road, Moconnell@okstate.edu ), firstname.lastname@example.org controls on mapping on aso b, Maggi Kelly b 31, Menlo Park, CA 94025, USA f Environment l sev ie r .com/ locate / rseNyman, Walters, Delaune, & Patrick, 2006).
Freshwater marshes, where emergent plant growth leads to peat formation, are found in some of the most significant coastal areas of the United States, including South Florida, the Mississippi River Delta, and the Sacramento–San Joaquin River Delta. The carbon stocks and 167K.B. Byrd et al. / Remote Sensing of Environment 149 (2014) 166–180future cumulative carbon storage of these marshes are referred to as “Blue Carbon,” and play an important role inmanaging atmospheric carbon (Pendleton et al., 2012). However freshwater marshes have often been modified by compaction, drainage and oxidation, nutrient enrichment and contamination (Deegan et al., 2012; Deverel & Leighton, 2010; Mishra et al., 2012; Nungesser, 2011; Tornqvist et al., 2008).
Marshes are also potentially impacted by sea level rise (National
Estuarine ResearchReserve System, 2012). If sea level rise is accelerated, coastal marshes characterized by low plant productivity and low sediment supply may experience shifts in the mix of intertidal habitats, leading to loss of vegetation and gains in low marsh and mudflats (Schile et al., 2014; Stralberg et al., 2011; Swanson et al., 2013). The loss or degradation of coastal wetlands could increase net global atmospheric CO2 inputs by ~6% per year (Hopkinson, Cai, & Hu, 2012). Therefore these ecosystems have been targeted for greenhouse gas (GHG) offset programs, markets and habitat restoration (Emmett-Mattox,