 |
Sap Flow Measurements and
Instruments |
|
|
Data and Measurement Theory
|
|
The heat dissipation method of Granier (1985, 1987) measures sap velocity as the temperature difference between a heated and an unheated probe inserted into the stem of a woody plant. The heated probe is powered from a constant-current (constant power) source and the difference in temperature is often measured using a thermocouple pair. Heat dissipation increases as water moves through the xylem, carrying heat away from the heated probe. Thus, as sap velocity increases, the temperature difference between the heated and non-heated probe decreases. |
 |
Granier tested his probes in three woody species as well as in other water-conducting materials. He determined that sap velocity ν (in mm s-1) was related to the temperature difference between probes (ΔT; Equation 1).
The maximum difference between probes is ΔTm when sap velocity (ν) is zero (Equation 2).
Granier also determined the coefficients in Equation 1 by fitting a nonlinear regression to the relationship between sap flow and difference in temperature. The rate of sap flow (in L s-1) is calculated in Equation 3, where the cross-sectional area of sap wood (the conducting xylem, as opposed to the heartwood) is A (in m2).
The difference in temperature can be predicted by rearranging the equations (and is used for estimating the sap velocity at any point along the probe; Equation 4). |
 |
(Eq.1) |
 |
(Eq.2) |
 |
(Eq.3) |
 |
(Eq.4) |
|
| After setting up your sap flow system and collecting the temperature difference data over a series of days, you should be ready to start analyzing the data. Hopefully you are also collecting micrometeorological data simultaneously, in order to give some context to these measurements. |
| Temperature difference data needs to be normalized by subtracting off the "zero flow" values. These values can be determined for every evening by hand, or automatically (we hope to have a program available soon). |
 |  |
| The data now must be transformed by the equations listed above, converting from the difference in temperature to velocity of water: |
using 0.119k1.232 to get mm s-1
or
using 0.000119k1.232 to get m3 H2O m-2 of sapwood area s-1 |
 |
|
The reason to use the last value is for ease of converting to volume flow if the sap wood area is known. For instance, for a 1 m diameter tree, if the outer 2 cm were conducting sap, then the sap wood area would be: |
Π x (outer radius2 − inner radius2)
3.1415 x (0.52 − 0.482) = 0.0615 m2
1 m3 H2O = 1000 liters
|  |
| Thus, multiply the velocity (in m3 H2O m-2 of sapwood area s-1) by 1000 and divide by 0.0615 to get liters s-1 of sap flow. Dividing further by distal leaf area should correlate well with transpiration on a leaf area basis (integrated over 24 h because of capacitance of trunks). |
One of the reasons to collect micrometeorological data concurrently with sap flow is to allow the interpretation of funny data. Specifically, the relative humidity, clouds (affecting radiative heating and cooling of leaves), and rain will strongly affect transpiration and thus sap flow measurements.
The example to the right, top panel, is the vapor pressure difference (VPD; the driving force for transpiration), calculated as the difference in water vapor concentration in the air (from relative humidity and air temperature or dewpoint) subtracted from the water vapor concentration inside of the leaf (assuming 100% water vapor saturation inside the leaf). On the days when the relative humidity was higher, the VPD was lower and the transpiration rate (as measure by sap flow, bottom panel) was significantly reduced. |
|