Optimal collocation of sensor-based atmospheric measurements and in-situ field observations

The National Ecological Observatory Network (NEON) is a continental-scale research platform with a projected lifetime of 30 years. NEON's purpose is to provide high-quality open source data that enable a better understanding and potential forecasting of environmental change (e.g., climate and land cover). To accomplish this, NEON is establishing 60 terrestrial sites, in which drivers and responses to environmental change are quantified. At each site, sensor-based measurements capture biophysical variables, such as radiation, temperature and evapotranspiration, and field technicians conduct in-situ observations of biotic variables, such as above- and below-ground biomass. To establish valid relationships between these observations of ecosystem drivers and responses, two contradicting requirements must be fulfilled: Both sensor-based and human-based observations shall be representative of the same ecosystem, while they shall not significantly interfere with one another.

Here we develop a procedure which quantitatively optimizes this trade-off through; (i) Determining the ratio of a user-defined impact threshold to effective impact area for different field technician activities. (ii) Quantifying the source area distributions of sensor-based measurements. (iii) Determining the range of feasible distances between sensor locations and technician activities by combining (i) and (ii). For a given field technician activity, the upwind distance from a sensor location required to stay below the same impact threshold differs among sites. These differences arise from site-specific environmental properties and corresponding differences in the sensor setups. As an example, we present results for three NEON sites in operations (located in FL, MA, and CO). For below- and above-ground biomass sampling, as well as insect traps, the minimum distances vary among sites in the range of 30 m–70 m, 35 m–105 m, and 180 m–245 m, respectively. Analogously, the maximum distance for mutual representativeness (90% cumulative flux footprint) varies between 290 m–610 m.

This strategy provides an evidence-based and repeatable method for combining sensor-based measurements and field observations at pre-defined levels of disturbance and spatial representativeness. The developed algorithm represents a general framework which is applicable to other environmental research sites where similar collocation is desired. Such a framework is an essential prerequisite to warrant establishing reliable relationships between ecosystem drivers and responses that are captured by different observation methods. Ultimately, the ability to improve our understanding of continental-scale ecology depends on the comparability of these relationships among research sites.