Abstract
ABSTRACT This paper describes how a validated semi-empirical, but physiologically based, remote sensing model – Ensemble_all – was up-scaled using MODIS land surface temperature data (MOD11C2), enhanced vegetation indices (MOD13C1) and land-cover data (MCD12C1) to produce a global terrestrial ecosystem respiration data set (Reco) for January 2001–December 2010. The temporal resolution of this data set is 1 month, the spatial resolution is 0.05°, and the range is from 55°S to 65°N and 180°W to 180°E (crop and natural vegetation mosaic is not included). After cross-validating our data set using in-situ observations as well as Reco outputs from an empirical variable_Q10 model, a LPJ_S1 process model and a machine learning method model, we found that our data set performed well in detecting both temporal and spatial patterns in Reco’s simulation in most ecosystems across the world. This data set can be found at http://www.dx.doi.org/10.11922/sciencedb.934.
Highlights
Terrestrial ecosystem respiration (Reco) is an important contributor to climate change (Le Quéré et al, 2009)
MODIS products with weekly 1 km resolution backed by FLUXNET data form an important approach for up-scaling site in situ observations to the global scale
Jägermeyr et al (2014) proposed a simple RECO model which is capable of up-scaling flux site observations to continental scale based only on MODIS enhanced vegetation indices (EVI) and land surface temperature (LST), but this is mostly an empirical method which lacks physiological basis, especially in the process of simulating reference respiration
Summary
Terrestrial ecosystem respiration (Reco) is an important contributor to climate change (Le Quéré et al, 2009). MODIS products with weekly 1 km resolution backed by FLUXNET data form an important approach for up-scaling site in situ observations to the global scale. Jägermeyr et al (2014) proposed a simple RECO model which is capable of up-scaling flux site observations to continental scale based only on MODIS enhanced vegetation indices (EVI) and land surface temperature (LST), but this is mostly an empirical method which lacks physiological basis, especially in the process of simulating reference respiration. Up-scaling field-measured respiration data using remote sensing information is an important way of enriching large-scale Reco data sets so that temporal and spatial patterns in the data can be cross-validated (Boulton, 2018). We conducted cross-validation of our data set using in-situ observations and the Reco outputs from the variable_Q10 model proposed by Yuan, Luo, and Li et al (2011), from the process model LPJ_S1 and from a machine learning data set proposed by Jung et al (2019) so as to: (1) see if our data set could detect the main temporal and spatial dynamics of Reco in all major ecosystems; and (2) find the regions where our data set and the other three data sets diverge and to explore the reasons behind this divergence
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
