The Influence of Instrumentation, Siting, Exposure Height, and Temporal Averaging Methodology on Meteorological Measurements from SJVAQS/AUSPEX

View More View Less
  • 1 Pacific Gas and Electric Company, San Ramon, California
© Get Permissions Rent on DeepDyve
Restricted access

Abstract

When designing field measurement networks, or when using the data from network archives, it is important to be aware of the influence that instrumentation, siting, data processing, and sensor exposure have upon the measurements. A unique opportunity arose to study this influence when in summer 1990, as part of the collaborative San Joaquin Valley Air Quality Study/Atmospheric Utility Signatures, Predictions and Experiments study (SJVAQS/AUSPEX), a vast meteorological measurement network was established for the central one-third of the state of California. Within the constraints of available resources, a variety of instruments, sites, site densities, sensor exposures, and data processing techniques were involved in the study. Based on the dataset from SJVAQS/AUSPEX, this paper was written to identify and quantify some common measurement influences for the benefit of users of the SJVAQS/AUSPEX data as well as the designers and users of other networks and datasets.

To characterize and quantify measurement influences, comparisons were made of data associated with different instrument types, sites, measurement heights, and temporal averaging methodologies. Comparative statistics were then developed and summarized. Little variability existed among the characteristics of surface instrumentation, but sounding instrumentation varied greatly in the degree to which atmospheric vertical structure was resolved. Vertical sampling resolutions ranged from 4.5 to 300 m, and winds were averaged over layer depths ranging from 15 to 600 m. The spatial location of measurements had a great influence on the measured direction of the wind. Hourly wind directions at neighboring surface sites differed by an average of 18° over the network with an rms variability of 55°. Upper-air values of these same statistics were 12° and 64°, respectively. A functional dependence of wind variability on site separation distance was apparent from the data. The influence of sensor exposure height on wind speed measurements, when comparing neighboring sites with different measurement heights, was less than expected. The average ratio in wind speed for such site pairs was only 1.18 (1.06 by day and 1.34 at night) for 10-m versus 2-m measurements. Temporal averaging methodology greatly influenced hourly wind values, especially for wind speeds less than 5 m s−1. Large differences were observed in hourly wind speed averages from scalar versus vector averaging, as well as in hourly wind direction averages from resultant versus unit vector averaging.

Abstract

When designing field measurement networks, or when using the data from network archives, it is important to be aware of the influence that instrumentation, siting, data processing, and sensor exposure have upon the measurements. A unique opportunity arose to study this influence when in summer 1990, as part of the collaborative San Joaquin Valley Air Quality Study/Atmospheric Utility Signatures, Predictions and Experiments study (SJVAQS/AUSPEX), a vast meteorological measurement network was established for the central one-third of the state of California. Within the constraints of available resources, a variety of instruments, sites, site densities, sensor exposures, and data processing techniques were involved in the study. Based on the dataset from SJVAQS/AUSPEX, this paper was written to identify and quantify some common measurement influences for the benefit of users of the SJVAQS/AUSPEX data as well as the designers and users of other networks and datasets.

To characterize and quantify measurement influences, comparisons were made of data associated with different instrument types, sites, measurement heights, and temporal averaging methodologies. Comparative statistics were then developed and summarized. Little variability existed among the characteristics of surface instrumentation, but sounding instrumentation varied greatly in the degree to which atmospheric vertical structure was resolved. Vertical sampling resolutions ranged from 4.5 to 300 m, and winds were averaged over layer depths ranging from 15 to 600 m. The spatial location of measurements had a great influence on the measured direction of the wind. Hourly wind directions at neighboring surface sites differed by an average of 18° over the network with an rms variability of 55°. Upper-air values of these same statistics were 12° and 64°, respectively. A functional dependence of wind variability on site separation distance was apparent from the data. The influence of sensor exposure height on wind speed measurements, when comparing neighboring sites with different measurement heights, was less than expected. The average ratio in wind speed for such site pairs was only 1.18 (1.06 by day and 1.34 at night) for 10-m versus 2-m measurements. Temporal averaging methodology greatly influenced hourly wind values, especially for wind speeds less than 5 m s−1. Large differences were observed in hourly wind speed averages from scalar versus vector averaging, as well as in hourly wind direction averages from resultant versus unit vector averaging.

Save