Hi, I want to estimate pasture biomass as reference data for remote sensing data. Can you give me recommandations for a rising plate meter? What is the measurement accuracy compared to destructive harvest? What is the price?
Hi,
you can find several rising plate meter retailers on the web, most of devices allowing at least automatic recording of measurements. Prices from some hundreds euros.
Since you can easily make hundreds of measurements in a given paddock, you would be probably get a better image of the biomass heterogeneity and possibly calibrate more precisely remote sensing data.
Be aware that
i) destructive harvests under the plate are required for establishment of grass height / biomass equation. Over 100 biomass/height pairs may be necessary for a good estimation of biomass from grass height.
ii) such equation may vary according to grass species and phenology, so the calibration process.
iii) soil surface micro-relief may increase significantly the residuals of this equation, and the size of the calibration data set
I agree with Maurice and add that the calibration is extremely important and difficult to do well, at least in a grazed pasture. The main issue being whether to cut to ground level, or some arbitrary grazing height. If ground level then the samples need to be washed and dried to remove soil. If to grazing level it very hard to standardise this amongst the quadrats being cut.
We calibrated our RPM meter each week in one trial. The regressions were "all over the place" and so we reverted to the recommendation of the manufacturer, or was it DairyNZ, to use the "supplied calibration" for the "season".
With two people working independently to estimate the mass in one 0.4 ha trial paddock, there was a wide discrepancy in the estimates, even though the calibration was the same for both RPMs.
The RPM probably gives a good paddock estimate of mass but I have doubts about how good it is when estimating mass in small plots used to calibrate remote sensing tools.
I did many small quadrat cuts to provide a proximal calibrate of a commercial NDVI sensor and couldn't get better than r2 of ~0.5 and so I really wonder how so many published data have r2 >0.8!
Good luck Holger
Hello Keith Betteridge,
Were your plots showing some soil? or was it vegetation alone?
NDVI is known to be saturated at leaf area index (LAI) ~2 (pending species).
Did you try a red-edge index?
- Badges
- Science topic
- Similar topics
- Geoscience
- Remote Sensing