Assembling a Sonde to Probe the Lower Atmosphere for Micrometeorological, Ecological, and Air-Quality Studies.

In this study we propose to use a prototype of an efficient, low-cost UAV based sensor system which is portable, scalable, and affordable Autonomous Measurement System (AMS). AMS is designed to measure in situ meteorological, biophysical, and surface parameters of a designated area with minimal user supervision, and it is designed to be easily reconfigurable, portable, and rapidly deployable. It can be used in several platforms, such as drones, tether balloons, or kites. The AMS has multiple sensor configuration options capable of estimating surface reflectance, albedo, remote sensing vegetation indices (e.g. normalize difference vegetation index, NDVI; and enhanced vegetation index EVI), and trace gases for air quality. To facilitate easy replication of AMS, readily available commercial off-the-shelf hardware is used with minimal customization.

The AMS control processor is a Raspberry Pi board (RPi). It is powered by a 2.3 Ah LiPo battery, which can give an autonomy of 1 hour or more, pending on loading and power consumption restrictions. The AMS has 4 configurations: basic, radiation, chemistry, and full. Sensors are measured using the RPi serial protocols, or using the 8 channel, 16 bits, analog-to-digital (AD) board. The basic configuration measures air temperature, relative humidity (Adafruit, DHT 22), pressure (Adafruit, BMP 180), gps (Adafruit, ultimate GPS) and a magnetometer (Adafruit, HMC5883L). Also, a 3D sonic anemometer has been added (Anemoment, Trisonica). The radiation configuration includes a Photosynthetic Active Radiation (PAR, Licor, model 190b), and spectroradiometer (SPEC, Ocean Optics, USB2000). The present chemistry configuration is capable to measure NO, NO2, O3, and isoprene (Alphasense, model A4, and PID) using the AD board channels. Data are stored and processed in the AMS, and some parameters are transmitted to a ground station computer a RF transceiver (Xbee, series 1 or Pro). The heaviest configuration has a weight of 0.956 Kg.

A careful calibration should be performed for the meteorological parameters and AD board. Figure 1 shows good agreement between the AMS sensors and quality-controlled instrumentation, and differences are due to offsets in the linear regression. Sonic anemometer has also been compared to a Campbell CSAT sonic anemometer and show a reasonable agreement with the horizontal components. Calibration for the trace gas sensors are ongoing. The sonde was used in a forest gap (figure 2). Profiles of temperature, humidity, and solar radiation show reasonable results, with a cooler, darker, and more humid understory than at the top of the canopy. Data from the spectroradiometer indicates that the difference between the infra-red and the visible wavelengths are bigger at the top of the canopy than the bottom. This leads to bigger NDVI values at the top of the canopy than at the bottom, ie, the top of the canopy is “greener” than at the bottom.