A comparison of sap flux density estimates using heat pulse velocity, thermal dissipation and heat field deformation methods

Background/Question/Methods

Transpiration in trees is commonly estimated by using heat to sense the rate of water movement in the stem xylem.  There exist several methods that have been developed to accomplish this, including Heat Pulse Velocity, Thermal Dissipation and Heat Field Deformation. We tested the performance of these three methods under controlled laboratory and uncontrolled field conditions.  Our objective was to determine whether there were differences in the accuracy of the three methods over a wide range of sap flux densities.  We used commercially available sensors of each type and tested them simultaneously at sap flux densities ranging from 5 to 80 cm³ cm^-2 h^-1, which is within the range of values reported in the literature.  The laboratory tests were conducted using 10 cut stem segments of American beech (Fagus grandifolia).  Flow rates were held constant by maintaining a constant head of water pressure on the cut stem segments.  The field comparison of the methods was conducted over a three day period by placing all three sensor types in a live American beech tree.

Results/Conclusions

Our results demonstrated that there were significant differences in the estimates of sap flux density from each of the three methods.  All sensors underestimated actual sap flux density. There were systematic errors in all three methods which increased with increasing sap flux density.  Overall, the most accurate method was Heat Pulse Velocity, but only after applying a published wound correction.  The least accurate was the Thermal Dissipation method, which is the most commonly used method at the present time.   There was an average error between actual and estimated sap flow density of 35% using Heat Pulse Velocity, 46% using Heat Field Deformation and 60% using Thermal Dissipation.  Although the Heat Pulse Velocity method was the most accurate, it had the highest noise level, especially at low flux densities.  The field test provided comparable results. The substantial improvement in the accuracy of the Heat Pulse Velocity method after using a wound correction suggests the Thermal Dissipation and Heat Field Deformation methods require similar methodological improvements.  At this time, calibrations are required for any of the methods to accurately estimate tree water use.  Each method has advantages and disadvantages for specific applications and these tradeoffs will also be discussed.