Probability Distributions for the Refractive Index Structure Parameter and the Inner Scale of Turbulence and Their Implications for Flux Averaging

• Main Authors: Andrea, Edgar L., Fairall, Christopher W., Persson, P. O., Guest, Peter S.
• Other Authors: ENGINEER RESEARCH AND DEVELOPMENT CENTER HANOVER NH COLD REGIONS RESEARCH AND ENGINEERING LAB
• Format: Text
• Language: English
• Published: 2003
• Subjects: Meteorology; Statistics and Probability; Fluid Mechanics; *PROBABILITY DISTRIBUTION FUNCTIONS; *TURBULENCE; *METEOROLOGY; RATIOS; FLUX(RATE); PATHS; REFRACTIVE INDEX; TIME SERIES ANALYSIS; BOUNDARY LAYER; SURFACES; SCALE; HYPOTHESES; BUDGETS; HEAT; BENEFITS; CONFIDENCE LIMITS; SCINTILLATION COUNTERS; BETA PARTICLES; ARCTIC OCEAN; Arctic; Arctic Ocean; Surface Heat Budget of the Arctic Ocean
• Online Access: http://www.dtic.mil/docs/citations/ADA421359; http://oai.dtic.mil/oai/oai?&verb=getRecord&metadataPrefix=html&identifier=ADA421359

Defining the averaging time required for measuring meaningful turbulence statistics is a central problem in boundary-layer meteorology. Path-averaging scintillation instruments are presumed to confer some time-averaging benefits when the objective is to measure surface fluxes, but that hypothesis has not been tested definitively. This study uses scintillometer measurements of the inner scale of turbulence l(sub 0) and the refractive index structure parameter (C(sup, sub n)) collected during SHEBA (the experiment to study the Surface Heat Budget of the Arctic Ocean) to investigate this question of required averaging time. The first conclusion is that the beta probability distribution is useful for representing; C(sup 2, sub n) and l(sub 0) measurements. Consequently, beta distributions are used to set confidence limits on C(sup 2, sub n) and l(sub 0) values obtained over various averaging periods. When the C(sup 2, sub n) and l(sub 0) time series are stationary, a short-term average of C(sup 2, sub n) or l (sub 0) can be as accurate as a long-term average. But, as with point measurements, when time series of path- averaged C(sup 2, sub n) or l (sub 0) values are nonstationary, turbulent surface fluxes inferred from these C(sup 2, sub n) and l (sub 0) values can be variable and uncertain-problems that path-averaging was presumed to mitigate. Since nonstationarty turns out to be a limiting condition, the last topic is quantifying the nonstationarty with a published nonstationarty ratio and also by simply counting zero-crossings in the time series.