Methods for wrangling and cleaning DEEP and environmental data
Valuable initiatives are currently collecting global databases of important tree physiological parameters, which could be utilized within a plethora of relevant forest ecological and tree physiological research activities.
The task at hand is to identify the “hotspots” of data platforms and the means to access and process the data, which is often not a straightforward task. Here we provide support for importing global, regional and site-specific data covering climatic, physiological and allometric measurements.
Moreover, interactive data formatting and curation procedures are presented which enhance data cleaning efficiency and reproducibility. The examples include data products provided in tabular Excel, txt, image-type / raster and netCDF formats.
Spatial and temporal analyses, hierarchical modelling
Statistical analysis is “at the heart” of performing hypothesis-driven research in the field of forest ecology and eco-physiology. With the ever increasing amount of available analyses techniques, clear guidance is needed to avoid common mistakes in data analysis and interpretation. Here we provide guidance on applying general statistical analyses on both spatial and temporally specific data. Moreover, uncertainty analysis and mixed-effect modelling frameworks are presented. We do not intend to explain basic statistical theory or model assumption testing, as more dedicated protocols and platforms exist, but rather highlight means of utilizing existing statistical tools in a dendro-ecological and eco-physiological context.
Techniques for typical DEEP data,
like tree rings, sap flow and dendrometer measurements
The range of user-specific choices (e.g., parameter values) applied during data processing between studies represents a key challenge for homogenising and synthesising time-series measurements, among others. Common techniques used within forest ecology and eco-physiology include; tree ring, sap flow, dendrometer, wood anatomical and eddy covariance data analysis. Proper usage of state-of-the-art techniques in processing such data can support robust conclusions about the state of the forest water and carbon cycle and how these are affected by climate change. Here we provide step-by-step examples of several data-processing techniques required for obtaining relevant and reproducible output variables.