Light use efficiency of peatlands: Variability and suitability for modeling ecosystem production
Peatland net ecosystem production is a key variable to assess changes in the functional role of peatlands in the global carbon cycle. Light use efficiency (LUE) models in combination with satellite data have been used to estimate production for most major ecosystems, but peatlands have been largely...
| Published in: | Remote Sensing of Environment |
|---|---|
| Main Authors: | , , |
| Format: | Article in Journal/Newspaper |
| Language: | English |
| Published: |
Elsevier
2016
|
| Subjects: | |
| Online Access: | https://lup.lub.lu.se/record/037afa5b-3f14-4f3b-8901-f68e4cdf4666 https://doi.org/10.1016/j.rse.2016.05.004 |
| Summary: | Peatland net ecosystem production is a key variable to assess changes in the functional role of peatlands in the global carbon cycle. Light use efficiency (LUE) models in combination with satellite data have been used to estimate production for most major ecosystems, but peatlands have been largely ignored. The objectives of this study were: 1) to examine how the LUE parameter epsilon, ε (the amount of carbon fixed or converted to biomass per unit absorbed photosynthetically active radiation), varies between and within four different peatlands; 2) to examine how the variations in ε relate to variations in environmental conditions; and 3) to evaluate a LUE-based model for estimation of ε in peatlands. We achieve these objectives using a combination of eddy covariance flux measurements, climate data and satellite data and estimate ε using the LUE-based vegetation photosynthesis model (VPM). The results show that: 1) mean site-specific flux-derived ε values (± standard deviation) were split into three statistically different groups: lowest values at the two colder fens, Kaamanen and Sandhill (0.22 ± 0.18 and 0.23 ± 0.20 g C MJ− 1, respectively), highest values at the treed fen La Biche (0.47 ± 0.27 g C MJ− 1) and intermediate values at the bog, Mer Bleue (0.34 ± 0.18 g C MJ− 1); 2) Variations in monthly ε within sites related mainly to air temperature, while variations in annual ε within sites related mainly to wetness variables; 3) relative mean absolute errors of estimates of ε for the four sites ranged between 19% and 35%, with r2 values ranging between 72% and 93%. LUE models are appealing as they are relatively simple formulations of variables that are easily obtained from satellite data. Challenges associated with the use of satellite data derived input variables are further discussed in the paper. |
|---|