proudly sponsored by
|
archiving of SOWACS proudly sponsored by |
|
22 Nov. 1996 Dear Sowacs and Trevor: I guess I opened a tangential discussion on micromet in this soil water forum. But I've learned quite a bit, and the exchanges have been very good. I would like to add some to Trevor's points and attempt to guide the discussion back to soil water. Michael Savage answered the temporal question so I can't add anything there. POINT 1 (point measurements) Trevor correctly pointed out the issues that all soil water balances are almost point measurements (including lysimeters that are just bigger points). I tried to emphasize earlier in discussion neutron scattering (NS) that point measurements of irrigation and rainfall (actually you need to know infiltration) are critical. This may be a REAL challenge for some irrigation methods as Trevor indicated. Atmsopheric methods like eddy correlation and Bowen ratio can't distinguish sources of water vapor on this fine detail. This is called "footprint" by micromet folks. The footprint for a BR/EC sensor may take several 100's of m. So the measurement is essentially integrating the measured vapor flux over this footprint. Implied in BR is that Rn and G are also "uniform" over this footprint, which may be more often valid assumptions. Of course, G has a more distinctly 2-D surface characteristic that must be integrated. POINT 2 (neutron probe calibrations) Trevor indicated that many neutron probe calibrations are linear involving only two factors -- a slope and an intercept. IF one calibration equation is satisfactory, then the intercept will not affect ET (the difference in 2 soil water measurements). HOWEVER, if a profile is not uniform and several calibration equations are needed to characterize this profile difference, then the intercepts may be more important factors in determing ET (still won't affect the layer in question though; but spatial variability of any layer is "unknown" to the user unless you poke a bunch of holes and ruin your site!). Terry Howell At 04:31 PM 11/22/96 +1100, Trevor Finch <rsne@mpx.com.au> wrote: >Calculation of ET rates using soil moisture measurements (water balance >method) depend on the assumption that there is essentially 2-dimensional, >vertical, flow. With point sources and sinks (drippers or vines, for >example), crop water use calculations need to be corrected for sensor >position relative to the source/sink. > >A sensor near a source (dripper), or mid-way between sources, may show >almost no change. > >Do Eddy correlation methods also have to take into account non-uniform flow ? > >On the question of calibration of soil moisture sensors (Terry A. Howell) it >is of interest that the purpose of the exercise is to quantify crop ET. >This is essential a *rate*, and the important calibration of the soil >moisture sensor is of course the *slope*, the intercept (assuming it is >linear) is not important. Of great importance for a sensor, however, is >repeatability from site-to-site, stability and hysteresis. > >In order to calculate actual daily ET rates (which is presumably what is >wanted), do the Eddy correlation methods have to integrate over a 24 hour >period ? > > >---- >Trevor Finch ********************************************************************* * Terry A. Howell, Ph.D., P.E. (806) 356-5746 * * USDA-ARS (806) 356-5750 (Fax) * * P.O. Drawer 10 tahowell@ag.gov (E-mail) * * Bushland, TX 79012 http://www.net.usda.gov/cprl/ (Internet) * * * * 1/2 mi. West I-40 South Access Rd. (shipping) * *********************************************************************