New Article: How High to Fly for Mapping ET?
We modeled actual evapotranspiration (ET) using drone imagery over a commercial potato field in the Central Sands region of Wisconsin.
While modeling ET using remote imagery has been used for decades, the use of these models in humid climates and using low elevation, high resolution imagery, is still relatively new. Drones have the benefit of being easier to schedule around problematic cloud cover that we frequently get in the Midwest and being able to fly at lower elevations. This brings up question of ‘how high to fly’? The lower you fly, the higher resolution you have, but then you also have higher battery usage, and longer flight and processing times. To test this, we flew at three heights (90, 60, and 30 m above ground level). We ended up not seeing a difference in the reliability of our modeled ET estimates between the difference elevations. This is also the first UAV-based model of actual ET in potato agroecosystems, which we compared to ET estimates from an onsite eddy covariance system.
Additionally, we ran uncertainty analysis on our model and found that the predominate source of uncertainty was in the spatial inputs for the model. This gives us a better understanding of the impact of best practices for imagery collection and where we need to improve. These findings most relevant in the context of a Midwestern US humid climate and annual specialty crop systems like potato, where plant-scale imagery may not result in additional management benefits to justify the cost, time, and data processing of very low-altitude imagery. This work was led by Logan Ebert as his first published dissertation chapter—congrats Logan!