As said in the preamble, the hiRes RS data is a unique opportunity to test the behavior of the meteorological models used in this study. It was thereby assumed that the physics in the numerical weather model represents well enough the truth that the radiosondes are assumed to provide. To test that hypothesis is beyond the scope of this work. The hiRes RS data was from UK sites only as this was the only region covered by the NWM. Figure 4.2 shows the locations of the five hiRes RS sites available in the UK. They are operated by the UK Met Office and the data were obtained via the British Atmospheric Data Center (BADC, http://badc.nerc.ac.uk/home/index.html). Information about their location is summarized in Table 4.11. To assess the performance of a model used in the ray tracing algorithm and to compare the NWM to RS in terms of meteorological data sources, exactly those coordinates were introduced in the ray tracing, so ex-act collocation is assumed. It was thereby assumed that exex-actly those coordinates were used in the NWM data assimilation scheme.
1The default resolution (2 decimal places for lat/long and nearest metre for height above mean sea level) of the coordinates is the one from the PC-CORA files. More precise coordinates are quoted when available from the WMO website.
400
500
700
850
950 surface or 1000 0
1000 2000 3000 4000 5000 6000 7000 8000
0 1000 2000 3000 4000 5000 6000 7000 8000
0 5 10 15 20 25
240 260 280 300
Water vapour pressure [mb]
Temperature [K]
Geopotential height [m]
Figure 4.1: Example of measured versus interpolated profiles. Red line is the
’observed’ water vapour partial pressure whereas the blue line is the interpolated water vapour pressure. The green and black lines respectively represent the ob-served and interpolated absolute temperature profiles. Data were taken from hiRes RS site Albemarle (UK) on July 10, 2005, at midnight. The labelled hor-izontal lines represent the equivalent NWM pressure levels. At those points, by construction, measured and modelled variables coincide.
To assess the meteorological models, only the regular profiles at 11 am and 11 pm are used, although sometimes those two profiles are completed with more launches to cover special events. At the time of writing, the data availability as provided on the BADC web site was not up to date, but it is included here for completeness. All data available from 2001 to 2007 (inclusive) were con-sidered. Table 4.2 provides the main figures on the data availability detailed
350˚
350˚
352˚
352˚
354˚
354˚
356˚
356˚
358˚
358˚
0˚
0˚
2˚
2˚
4˚
4˚
50˚ 50˚
52˚ 52˚
54˚ 54˚
56˚ 56˚
58˚ 58˚
60˚ 60˚
0 100 200
km
Albemarle
Camborne
Herstmonceux Lerwick
Watnall
Figure 4.2: The five hiRes RS sites in the UK.
ing 2001, the data available were used. Generally speaking, the availability is very good. There are missing data but not enough to significantly affect the results presented here. The statistics given in Table 4.2 only consider launches that took place at either 1100 or 2300 UTC so the numbers might be different from the ones published on the BADC web site (http://badc.nerc.ac.uk/data/rad-highres/status.html) where every launch is accounted for. Since April 2002, all the RS sites appear to have launches twice a day only, whereas up to four launches were performed before that date (assuming all are available from BADC).
As the statistics on the availability here are only given for hours 11 and 23 UTC of each days, there are approximatively 60 launches a month. This al-lowed the comparison to be consistent over the full period considered for the NWM ray tracing and GPS data processing. Table 4.3 provides statistics on
the accuracy and precision of the RS measurements. The raw data are in the PC-CORA format. For a full description of the format see Vaisala [1998] or visit http://badc.nerc.ac.uk/data/rad-highres/pc-coradata.html for gen-eral information. Only the information of the met variables involved in the present work are reproduced here in Table 4.3.
To represent the humidity, it had to be decided wether to interpolate directly on e computed from the relative humidity on each pressure layer, or on a more ad-equate meteorological quantity such as the relative humidity (RH) or the specific humidity (q) for example, and then convert the interpolated value to e, needed for the computation of the wet refractivity and density. The determination of the humidity field is of prime interest as it is known to be the most difficult to represent. So error in the modelling itself should be kept minimal. RS reports do not usually contain the measured RH but instead the derived dew point temper-ature Tdew. This could be due to the fact that Tdew is used in the production of tephigrams.
Table 4.1: Coordinates of the five hiRes RS sites in the UK.
Radiosonde site ID Latitude Longitude Orthometric Geoid Ellipsoid [deg] [deg] height [m] height [m] height [m]
Albemarle albe 52.02 −1.88 141 50.574 191.574
Camborne camb 50.218 −5.329 88 53.344 141.344
Herstmonceux hers 50.89 0.319 54 45.052 99.052
Lerwick lerw 60.139 −1.183 82 49.665 131.665
Watnall watn 53.005 −1.250 117 49.201 166.201
Table 4.2: Monthly data availability for the five hiRes RS sites in the UK for the period January 2001 to March 2007 inclusive.
Potential Albermale Camborne Herstmonceux Lerwick Watnall
Year Month number of Nb. of Avail. Nb. of Avail. Nb. of Avail. Nb. of Avail. Nb. of Avail.
profiles profiles % profiles % profiles % profiles % profiles %
2001
TOTAL 4562 3429 75 3735 81 3414 74 4475 98 4125 90
Fortunately, the binary data from the digicora equipment does contain the mea-sured RH so allowed for a detailed comparison between raw meamea-sured RH values and the ones computed from the usually reported T and Tdew values. The follow-ing section (4.3) investigates the level of agreement between the measured and the reported (or computed or backed up) RH value, as this directly influences the wet and total refractivities determination.
Table 4.3: Published range of operation, resolution, accuracy and lag of RS me-teorological measurements.
Range of operation Resolution Accuracy Lag
Pressure [hPa] 3.0 – 1060.0 0.1 0.5
-Temperature [◦C] -90.0 – 60.0 0.1 0.2 < 2.5 s
Relative humidity [%] 0 – 100 1.0 2.0 < 1.0 s