Process-based Models as a Gap Filling Method for Eddy Covariance Measurements of Net Ecosystem CO2 Exchange: A Case Study for the Perennial Grass Miscanthus
Abstract
The eddy covariance (EC) method has been used to quantify the carbon fluxes between terrestrial ecosystems and the atmosphere for various types of ecosystems all over the world. Since the EC method provides ~108 measurements per year, inevitably there will be missing data for short- and long-term measurement periods due to equipment failure, system maintenance shutdowns, power-failures, unfavorable meteorological conditions and data filtering. Therefore, it is important to define the robust strategies for the gap-filling EC data. There are a wide range of gap filling approaches in the literature, predominately including linear and non-linear regressions, look-up tables, artificial neural networks. However, such procedures may be site-specific and limited in terms of insight into the underlying biophysical processes that drive fluxes. In process-based models, the flux prediction is based not only on driving variables such as meteorological variables but also on state variables such as soil type, soil texture, soil depth, and canopy height. Typically gap-filled EC data are used to validate process-based models, however at sites with long term flux data (e.g. 10 years) it may be possible to calibrate a processed based model on a subset of the data and then use the calibrated model to fill in data gaps. Agro-IBIS is a process-based model which provides hourly CO2 flux prediction for several annual and perennial cropping systems. The University of Illinois Energy Farm is a long-term measurement (over 10 years) of CO2 flux using EC method for perennial biofuel crops. In this study, the Agro-IBIS model was applied for miscanthus perennial biofuel crop to predict hourly CO2 fluxes for 10 years (2008-2017). To evaluate the Agro-IBIS model as a gap filling method, non-gap EC flux data subsets were compared with Agro-IBIS results. The primarily results showed that Agro-IBIS CO2 flux prediction has a good correlation (~ % 70) with EC flux data, and was similar to the more commonly used approaches. This suggests that Agro-IBIS can be used as a gap filling approach. Future work will include testing this approach at a similar EC miscanthus site that was recently established at Iowa State University.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2019
- Bibcode:
- 2019AGUFM.B41K2499A
- Keywords:
-
- 3315 Data assimilation;
- ATMOSPHERIC PROCESSES;
- 0414 Biogeochemical cycles;
- processes;
- and modeling;
- BIOGEOSCIENCES;
- 0426 Biosphere/atmosphere interactions;
- BIOGEOSCIENCES;
- 0430 Computational methods and data processing;
- BIOGEOSCIENCES