high marsh

Eddy flux measurements during 2012 from high marsh (Spartina patens/short Spartina alterniflora) off Nelson Island Creek, Rowley, Massachusetts


We deployed an eddy covariance system to measure ecosystem-atmosphere exchange of CO2 above a high marsh system (Spartina patens, short Spartina alterniflora) located on the Parker River Wildlife Refuge in marshes of Plum Island Sound, Rowley MA. This data represents the growing season CO2 exchange (May-October) in 2012.

Core Areas: 

Data set ID: 



Short name: 




Data sources: 



We deployed one eddy covariance system to measure NEE of the dominant high marsh at PIE. The location of the tower is ca. 2km away from the parking lot at Stackyard Road within the Nelson Creek catchment. Deployment of the system is seasonal from mid-April to mid-November. The site was equipped with a Campbell Scientific® Closed Path System (CPEC200). Micrometeorological instrumentation was mounted on a tower at a height of 4.16m above the marsh surface at low tide. 10Hz rawdata were stored on a 2 Gb PCMCIA card and downloaded every two to three weeks. Power for the datalogger and instrumentation was located ca.8 m to the northeast of the tower and consisted of four south-facing solar panels and four 6 V 225 amp-hour batteries.

Environmental data were recorded as 10min averages. Air temperature and relative humidity were monitored at the same height as the anemometer (Campbell Scientific HC2S3 enclosed in a naturally aspirated radiation shield). A four-component net radiometer (Hukseflux NR01) was mounted 1.5 m aboveground of the high marsh. At the same height, two sensors (LI190SB, Licor) monitored incoming and reflected photosynthetically active radiation (PAR). In addition, a pressure transducer (Campbell Scientific CS456) recorded water table height at the high marsh. Soil temperature at a depth of 2cm, 6cm, 10cm, 20cm and 40cm was measured with (TCAV-L; Campbell Scientific; Logan, Utah, USA), and soil heat flux at a depth of 8 cm was measured with two soil heat flux plates (HFP01-SC; Campbell Scientific; Logan, Utah, USA). This data was recorded on a separate CR3000 datalogger.

Turbulent fluxes of momentum, sensible heat, latent heat and CO2 were determined by the eddy covariance technique (Baldocchi et al. 1988). Half hourly CO2 and H2O fluxes were calculated as the covariance between the turbulent departures from the mean of the 10 Hz vertical wind speed measured with a 3D sonic anemometer (CSAT3; Campbell Scientific; Logan, Utah, USA) and the CO2 and H2O dry mixing ratio measured with the closed path analyzer. Fluxes were processed using EdiRe software (Robert Clement, University of Edinburgh) and reported using the meteorological sign convention where negative NEE indicates carbon uptake and positive NEE indicates carbon loss from the ecosystem. Two coordinate rotations were performed on the wind components, and the time lag between wind and CO2 mixing ratio measurements was determined and removed for each averaging interval of 30min. For every 30 min period, a factor for the correction of the frequency attenuation of the flux was calculated according to Moore [1986] and applied to the flux. Fluxes were calculated using the Edire software (version, R. Clement, University of Edinburgh, UK). Afterward, fluxes were filtered for system malfunctioning and calibration periods, integral turbulence characteristics, stationarity, and wind direction [Foken etal., 2012]. We also excluded measurements when less than 75% of the flux was generated within the study area. Thresholds in friction velocity (u∗) for nighttime fluxes were determined according to Papale et al. [2006] and set to 0.14m/s .

To continuously monitor aboveground biomass, we calculated a broadband normalized difference vegetation index (NDVI) based on the approach of Wilson and Meyers [2007]. Incoming (i) and reflected (r) Solar (S) and photosynthetically active radiation (PAR) measurements were converted into red and near-infrared reflectance. Solar zenith effects were removed by using data exclusively around solar noon (10 A.M.–2 P.M. EST). In our system, spring tides occurred around noon, so that simultaneous radiation measurements recorded the effect of tidal inundation at that time. A decrease in NDVI would reflect that during inundation the amount of biomass that was air exposed was smaller than under nonflooded conditions. We included this effect in our NEE model by creating two continuous time series of NDVI to simulate flooded and nonflooded conditions: NDVIall which included spring tide effects, and a reference time series, NDVIref, which represented nonflooded conditions.

NEE is gap-filled with a modified PLIRTLE model (NEE=GPP+Reco), using NDVI_all, air temperature and PAR as input. GPP_all and Reco_all are estimated using the two sub-models of the PLIRTLE model. GPP_ref and Reco_ref are modelled with NDVI_ref as input variable and thus represent the fluxes occurring if no tidal inundation had occurred.

Baldocchi et al. [1988]: Measuring Biosphere-Atmosphere Exchanges of Biologically Related Gases with Micrometeorological Methods. Ecology, Vol. 69, No. 5, pp. 1331-1340.

Moore [1986]: Frequency response correction s for eddy correlation systems. Boundary Layer Meteorology, Vol. 37, pp. 17-35.

Foken et al. [2012]: Corrections and Data Quality Control. In: Aubinet, Vesala, Papale (editors): Eddy covariance - a practical guide to measurement and data analysis.

Papale et al. [2006]: Towards a standardized processing of Net Ecosystem Exchange measured with eddy covariance technique: algorithms and uncertainty estimation . Biogeosciences, Vol. 3, pp. 571-583.

Forbrich, I., and A. E. Giblin (2015), Marsh-atmosphere CO2 exchange in a New England salt marsh, J. Geophys. Res. Biogeosci., 120, doi:10.1002/2015JG003044.


Data collection and processing complete

Version 01: January 21, 2016, data and metadata created to comply with importation to Drupal and LTER PASTA. Used MarcrosExportEML_HTML (working)pie_excel2007_Jan2015.xlsm 1/15/15 4:26 PM for QA/QC to EML 2.1.0


Subscribe to RSS - high marsh