ABSTRACT:

Soil carbon is the largest active terrestrial reservoir in the carbon cycle, and potential feedbacks involving soil carbon play an important role in future climate change. Understanding how combinations of factors, such as vegetation, temperature, and soil moisture, affect soil carbon dioxide (CO2) production in various environments across sub-daily to seasonal timescales is essential to accurately predict climate impacts on the carbon cycle. However, in-situ high resolution data are rare and often measure surface fluxes instead of deeper soil fluxes. Here we present a quantitative accounting of factors governing CO2 production in agricultural and prairie soils, using high-resolution monitoring of below-ground soil CO2 concentrations and estimates of soil respiration fluxes. Across sites, we find that Normalized Difference Vegetation Index (NDVI) tends to predict soil CO2 concentration and production more effectively than soil temperature at daily timescales. At the time scale of a rain event, rain frequently leads to rapid drops in CO2 concentration due to soil CO2 abiotically equilibrating with rain water followed by prolonged increases in inferred CO2 production. This pattern was only visible due to the high resolution of the soil CO2 concentration data. We also found that prairie soils, which host a greater diversity of plant species, have a higher rate of CO2 production than agricultural soils under comparable climate drivers. Finally, we examine how the temporal resolution of soil CO2 data affects the magnitude of environmental correlations. These findings highlight that seasonal environmental and vegetation conditions strongly influence local soil CO2 responses.

ABSTRACT:

data for the paper A low power low cost chamber based CO2 sensor needed to run the code attached to the paper

ABSTRACT:

This dataset includes four streams with daily CO2 data at five sites labeled daily data. There is an additional data set labeled LTM sites that has daily CO2 and stream chemistry at 4 sites. Finally, there is a file labeled reach characteristics that has the number of seeps, the slope, and other reach related measurements.

ABSTRACT:

Saccardi and Winnick Data contains the East River sample data including chemistry and corrected pCO2 data (n=162) with the included column denoting whether data points were represented by NHDplus data point (n=121) and therefore used in the model.