Indian Journal of Radio & Space Physics

Vol 40, December 2011, pp 301-310

Monitoring tropospheric radio refractivity over Nigeria Using CM-SAF data

derived from NOAA-15, 16 and 18 Satellites
Babatunde Adeyemi$,* & Israel Emmanuel
Department of Physics, Federal University of Technology, PMB 704, Akure, Ondo State, Nigeria
$
E-mail: tundebx@yahoo.com

Received 25 April 2011; revised 23 September 2011; accepted 11 October 2011

The variability and structure of tropospheric refractivity (N) over Nigeria has been studied using monthly data of
pressure, temperature and relative humidity for the period 2004-2006 for twenty six stations, obtained from the archives of
the Department of Satellite Applications Facility on Climate Monitoring (CM-SAF) DWD, Germany. The data was
retrieved from ATOVS observations onboard polar orbiting NOAA-15, 16 and 18 satellites. The results have shown that
variations in each region and at different atmospheric levels are influenced by the North–South movement of inter–tropical
discontinuity (ITD). Using the analysis of variance (ANOVA) technique on the climatological data, an empirical
relationship of the form N = ω + γNs (ω, γ constants) have been established between columnar refractivity and its surface
value, Ns, for the four regions (coastal, Guinea Savannah, midland and sub-Sahelian) in Nigeria. The initial tests carried out
on the model using data obtained over the regions for the year 2007 yielded an encouraging result as have been established
by the use of Kolmogorov-Smirnov tests.

Keywords: Inter-tropical discontinuity (ITD), Columnar refractivity, Tropospheric refractivity; Surface refractivity,
Analysis of variance (ANOVA)

PACS No.: 92.60.hf

1 Introduction knowledge of the refractivity variations is very

A reliable operation of ground-based essential so as to be able to design reliable and
communication systems for various purposes largely efficient radio communication (terrestrial and
depends on the physical state of the atmospheric satellite) systems.
boundary layer. Meteorological situations of the The refractive index in the troposphere falls slowly
troposphere in radio paths are governed by various with height and the resulting refraction causes the
properties of the under surface layer, its physical and radio horizon to appear to be 1.33 times further away
geographical as well as climatic features of the region. than the geometric horizon9. Korak10 in his study on
Changes in atmospheric conditions and its consequent “neutral atmospheric refraction on microwave
effects on tropospheric radio wave propagation are propagation and its implication on GPS based ranging
manifested in the variations of the radio field systems” pointed out that the tropospheric refraction
strengths and the distance to the radio horizons1-3. affects microwave propagation by retarding and
Microwave propagation through the troposphere is bending it, causing an error in microwave ranging.
affected by varieties of natural phenomena caused by Adediji & Ajewole11, in their study on vertical profile
the presence of some meteorological parameters4-7. of radio refractivity gradient in Akure, Nigeria, noted
These meteorological parameters, viz. pressure, that one reason for multipath of electromagnetic wave
temperature, and relative humidity have serious effect is bending due to variation in the refractive index
on radio wave propagation at UHF and microwave distribution along layers of the atmosphere.
frequencies. These effects are analyzed from the study
of radio refractive index derived from these The atmospheric refractive index is slightly greater
meteorological parameters8. Since these than unity (about 1.0003), hence the refractive index
meteorological parameters vary considerably of air is measured by a quantity called the radio
diurnally and seasonally in the tropics, quantitative refractivity, N, defined as12,13:
302 INDIAN J RADIO & SPACE PHYS, DECEMBER 2011

n = 1 + N × 10−6 …(1) 2 Source of data and Processing techniques

The source of upper air climatological data used for
present analyses is the Department of Satellite
N = ( n − 1) × 10 6 …(2)
Application Facility on Climate Monitoring (CM-
SAF), DWD, Germany. The CM-SAF focuses on the
Studies by Adeyemi & Adedayo14 on atmospheric atmospheric part of the essential climatic variables
radio refractivity and water vapour density at Oshodi defined within the framework of the Global Climate
and Kano, Nigeria, revealed that atmospheric radio Observing System (GCOS). The CM-SAF
refractivity is generally high during the rainy season operationally applies the international ATOVS
at all the levels of the atmosphere considered while its processing package (IAPP) to retrieve humidity and
values fall during the Harmattan period. Bean & temperature from ATOVS observations onboard
Thayers15 noted that correlation exists between the NOAA-15, 16 and 18. The ATOVS flies on NOAA
monthly means of surface refractivity, Ns, and polar orbiting satellites, and is composed of Advanced
monthly means of the refractivity decrease in the first Microwave Sounding Unit (AMSU) and High
kilometre above the sea level. Bayong & Resolution Infrared Radiation Sounder (HIRS/3)
Djakawinata16, while trying to study the influence of (HIRS/4 on NOAA-18 and MetOp). The AMSU
meteorological factors on tropospheric refractive platform is composed of two separate radiometers,
index over Indonesia using radiosonde data, noted namely AMSU-A and AMSU-B (MHS on NOAA-18
that the value of the index of refraction ‘n’ near the and MetOp). AMSU-A and –B are cross-track
surface in the maritime continent of Indonesia is scanning total power radiometers with instantaneous
about 1.000378 (N=378) in the rainy season of fields of view (FOV) of 3.3° and 1.1° providing
January while the value of the refractive index is nominal spatial resolutions at nadir of 48 km and
about 1.000368 (N=368) in the transition period of 16 km, respectively. The 15 AMSU-A channels
October. They concluded that the difference observed primarily provide temperature sounding of the
in the refractive index values is caused by the atmosphere through channels located in the
difference in the water vapour content in the 57 GHz O2 absorption band. It also has three channels
atmosphere in both the seasons. (23.8, 31.4, 89 GHz) that provide information on
Stergios et al.17 observed that from the refractivity tropospheric water vapour, precipitation over ocean,
variations of the Helenic troposphere, radio sea ice coverage, and other surface characteristics.
refractivity increases from January to June where it AMSU-B has five channels that mainly measure
appears to reach its peak, and decreases from July to water vapour and precipitation over land and sea.
its minimum in August. Its frequencies at 183.31±1.00, 183.31±3.00 and
Fairall18, in his study on a top down and bottom-up 183.31±7.00 GHz are positioned within a water
diffusion model of CT2 and CQ2 in the entraining vapour absorption band, and the 89.0 and 150 GHz
convective boundary layer, revealed that within the channels are placed within atmospheric windows. The
atmospheric boundary layer (ABL), the atmospheric International ATOVS Processing Package (IAPP) ver
refractive index fluctuates chaotically in time and in 7 is used to retrieve profiles of atmospheric
all three spatial directions and near the top of the temperature and mixing ratio as well as other different
convective boundary layer, the values of the structure atmospheric parameters at global scales. The
function parameters (which include temperature, validation of water vapour and temperature products
water vapour and relative humidity) typically depart is carried out using global radiosonde observations
from those of the mixed layer forms and increase that meet the quality standards of GCOS Upper Air
sharply to high values before decreasing again at Network (GUAN). The ATOVS retrieved temperature
greater heights. and mixing ratio profiles have been found not to be
The objective of this paper is to study the space – independent from model fields from the numerical
time distribution of tropospheric radio refractivity weather prediction model taking into consideration
over Nigeria using data retrieved from NOAA-15, 16 the biases and the bias corrected root mean square
and 18 satellites and to propose an empirical model error (RMSE). The profiles are vertically integrated
relating surface radio refractivity, NS, to radio and averaged to provide temperature and humidity for
refractivity aloft for four different regions in Nigeria. 5 layers (925, 775, 600, 400 and 250 mb). The data
 ADEYEMI & EMMANUEL: MONITORING TROPOSPHERIC RADIO REFRACTIVITY OVER NIGERIA 303

obtained for the period 2004-2006 was used in Meteorological factors (pressure, P; air
computations of tropospheric refractivity for twenty temperature, T; and relative humidity, H) are obtained
six stations in Nigeria (Fig. 1), classified into four directly from the retrieved data. Water vapour
regions based on their climatic conditions19. pressure, e, on the other hand, was determined using23
As in Adeyemi13, Aro20 and Adedokun21,22, radio
refractivity between surface and 925 mb level has HeS
been labelled as low-level refractivity (NL); between e= …(6)
775 mbar and 600 mbar as mid-level refractivity 100
(Nm); and between 400 mbar and 250 mbar as upper-
where, H, is relative humidity (%); es, saturation
level refractivity (Nu).
vapour pressure (hPa). It should be noted that Nwet is
2.1 Data analysis techniques mainly responsible for the variability in N within the
The radio refractivity, N, of air for frequencies up troposphere24,25. Nwet contributes about 34% to the
to 100 GHz is generally expressed as8,13: total value of Ns while Ndry contributes about 66%.
Radio refractivity was determined using Eq. (4) for
77.6  4810 e  the five different layers.
N= P +  …(3) Then for each region, correlations of (NS and NL),
T  T 
(NS and Nm) and (NS and Nu) were obtained. High
P e correlation values were found to be predominant at
N = 77.6 + 3.75 × 105 …(4) most of the regions, most especially, at the lower and
T T2 middle levels of the atmosphere (Table 1), therefore,
using analysis of variance (ANOVA) technique (table
N dry and N wet term components from Eq. (4) are
not shown), a linear regression relation of the form
defined as:
N = ω + γN S …(7)
P e
N dry = 77.6 ; N wet = 3.75 × 105 …(5)
T T2 was developed taking into consideration the
associated F-ratios and the probability values.
where, P, is the atmospheric pressure in hPa; T, air
temperature in Kelvin; and e, water vapour pressure in 3 Result and Discussion
hPa. 3.1 Space – time distribution of tropospheric radio refractivity
In the coastal region, Figs 2(a-d), both the surface
refractivity, NS, and low-level refractivity, NL, display
double peaks with a dip in between them around
July/August. Their mean values during the dry season
are 373.3±2.96 and 327.2±2.57, respectively, while
during the rainy season they were 374.7±1.37 and
328.4±1.19, respectively. The mid-level refractivity,
Nm, on the other hand displays seemingly double
peaks with the first in May/June and the other in
September. The observed double peaks here are not as
conspicuous as in NL. Nm values are generally lower
than those of NL. This is expected to be so as water
vapour decreases with increasing height from the
surface. Its average value during the dry season is
217.5±0.95 while during the rainy season, it is
222.9±0.26. Upper level refractivity, Nu, has its values
slightly increasing from January to July, after which,
it then decreases for the rest of the year. It lies
between 109.4±0.085 and 110.3±0.095, with its mean
Fig. 1 — Map of Nigeria showing study locations value being 109.8±0.09.
304 INDIAN J RADIO & SPACE PHYS, DECEMBER 2011

Table 1 — Values of the best fit parameters ω and γ in the regression equation of NL on NS, Nm on NS and Nu on NS
(a) Low level k NL + ω + γNs
Region Ω Error in ω Γ Error in γ r CD, % F-ratio p-value

Coastal 12 3.285 0.005 0.868 1.22E-05 0.99 98.0 5.05E+09 0.000

Guinea Savanna 12 3.231 0.012 0.868 3.27E-05 0.99 98.0 7.05E+08 0.000
Midland 12 89.02 0.409 0.688 0.001 0.99 98.0 3.02E+05 0.000
Sub - Sahelian 12 95.73 0.606 0.691 0.002 0.99 98.0 1.17E+05 0.000
(b) Mid level k Nm + ω + γNs
Region Ω Error in ω Γ Error in γ r CD, % F-ratio p-value

Coastal 12 173.7 82.16 0.126 0.219 0.18 31.9 0.329 0.060

Guinea Savanna 12 77.21 31.46 0.383 0.084 0.82 67.5 20.76 0.034
Midland 12 149.8 4.874 0.209 0.015 0.96 95.2 196.3 0.000
Sub - Sahelian 12 161.7 3.702 0.173 0.012 0.98 95.2 196.9 0.000
(c) Upper level k Nu + ω + γNs
Region Ω Error in ω Γ Error in γ r CD, % F-ratio p-value

Coastal 12 124.8 7.875 -0.040 0.021 -0.52 26.7 3.643 0.000

Guinea Savanna 12 109.5 3.166 0.001 0.009 0.03 0.00 20.76 0.034
Midland 12 108.5 0.722 0.004 0.002 0.50 24.7 3.286 0.000
Sub - Sahelian 12 109.6 0.405 0.001 0.001 0.17 2.91 196.9 0.000

Fig. 2 — Monthly mean variations of: (a) surface refractivity, Ns; (b) low level refractivity, NL; (c) mid level refractivity, Nm; (d) upper
level refractivity, Nu in the coastal region
 ADEYEMI & EMMANUEL: MONITORING TROPOSPHERIC RADIO REFRACTIVITY OVER NIGERIA 305

In the Guinea Savannah region, average NS and values of 294.5±5.19 and 291.5±3.64, respectively.
average low-level refractivity display similar pattern During the rainy months (April – October), high
with those of the coastal region [Figs 3(a-d)]. Here values of NS and NL are discernible. Their mean values
double peaks with a dip in between them in June- are 349.5±3.26 and 329.3±2.20, respectively. Nm
August are also discernible. The June-August dip is displays a single peak during May and August with
characterized by a slight increase in July. Their mean low values prevailing during the dry season and high
values during the dry season are 370.4±4.48 and values characterizing the rainy season. Mean Nm
324.7±3.89, respectively. During the rainy season, values during the dry and rainy periods are
they are 376.2±3.32 and 329.7±2.88, respectively. Nm, 211.0±0.65 and 222.4±1.13, respectively. In the case
on the other hand, display double peaks with a dip in of Nu, its values gradually decrease from January to
between them lasting between June and August. This April. The month of April seems to be where the
depression is more conspicuous than in Nm for coastal minimum for Nu occurs. A gently increasing trend
region. Its mean value is 219.7±1.09. Upper-level was then noticeable between April and July. July is
refractivity, Nu, produces a single peak in June. The the month where the annual peak occurred. Its mean
values are uniformly low, and lie between 109.4± value is 109.819 ± 0.0642 and its range lies between
0.05 and 110.2± 0.06 throughout the year. They gave 109.4±0.0643 and 110.2±0.0641 showing that in both
a mean value of 109.9± 0.065 the coastal and guinea savannah regions, Nu is
In the midland region, the structure of NS and NL uniformly low.
are well related to those of the Guinea savannah In the sub-Sahelian region, the variations observed
region [Figs 4(a-d)]. They are comparably low during in the refractivity parameters are almost similar to
the dry months (November – March), with mean those of the midland region [Fig 5(a-d)]. NS and NL

Fig. 3 — Monthly mean variations of: (a) surface refractivity, Ns; (b) low level refractivity, NL; (c) mid level refractivity, Nm; (d) upper
level refractivity, Nu at Guinea Savanna region
306 INDIAN J RADIO & SPACE PHYS, DECEMBER 2011

are low during the dry months with mean two seasons. They are the Harmattan and the rainy
refractivities of 251.7±3.47 and 269.4±2.52, season periods. During the Harmattan period, the
respectively. Their values are high during the rainy Nigeria troposphere is characterized by dry dust
season with mean refractivities of 327.2±11.57 and particles transported from Azores’s sub-tropical high
321.9±7.90, respectively. The variations observed in pressure system in the Sahara desert by the north-
Nm over this region resemble those of NS and NL. But easterly tropical continental (cT) air mass blowing
the mean refractivity values of Nm are lower than inland. This coincides with the period of no rain. The
those of NS and NL; they are 205.2±1.56 and rainy season, on the other hand, is characterized by
218.2±2.29 during the dry and rainy seasons, high humidity (heavy rainfall) brought about by the
respectively. Nu, on the other hand, has almost south-westerly tropical maritime (mT) air originating
uniform values throughout the year. However, its from the southern hemisphere. Wedged in between
value gently decreases during the late dry season the air masses, is a front known as inter-tropical
(January – April) before it rises, gently, to its highest discontinuity (ITD) (refs 26,27). The ITD migrates in
value in December. Its mean value is 109.9 ± 0.064 north-south and south-north direction reaching its
and it lies between 109.3±0.065 and 110.1±0.063. maximum northward extent 22-25°N in August and
From the foregoing, the observed variations in the its southward extent 4-6°N in January. While
refractivity parameters are attributable to the migrating, the ITD oscillates backward and forward
prevailing weather conditions pervading the country within a few latitudinal points. In January, all regions
(ref. 3). The Nigeria troposphere is characterized by north of 4°N are under the influence of the cT air.

Fig. 4 — Monthly mean variations of: (a) surface refractivity, Ns; (b) low level refractivity, NL; (c) mid level refractivity, Nm; (d) upper
level refractivity, Nu at the midland region
 ADEYEMI & EMMANUEL: MONITORING TROPOSPHERIC RADIO REFRACTIVITY OVER NIGERIA 307

Conditions are such that little or no precipitation is The values of the parameters ω and γ for NL, Nm, and
experienced and refractivity parameter values are low, Nu together with the correlation coefficient, r,
and show marked decrease from the coast inland. coefficient of determination, CD and probability
During March and April, the ITD position would be (p–value) at which the null hypothesis was either
between 10°N and 15°N meaning that the regions are accepted or rejected are as shown in Table 1(a-c). The
now beginning to experience little amount of rainfall degree of association evident in Table 1 shows that NS
making variability in the refractivity parameters to be is better associated with low-level refractivity (NL)
higher. On reaching its most northern position in than those of Nm and Nu at all the regions. The degree
August, all areas in Nigeria would have been of association between NS and Nm has also been found
subjected to widespread rainfall and increased cloud to be higher at both the midland and sub-Sahelian
cover. Nm and Nu values increase and become less regions than at the coastal and Guinea Savannah
variable. NS and NL on the other hand, in the coastal regions. Regression relations obtained for all the
and guinea savannah regions, experience lower values refractivity parameters are statistically significant at
than they are in the midland and sub-sahelian regions. all the levels and at the regions except Nm in the
This is as a result of the existence of the little dry coastal region which is statistically insignificant
season that often pervades the area around the (p-value = 0.060).
West-African coast in August. To verify the reliability of the linear relations
obtained in Table 1, they were then applied to
3.2 Regression analysis evaluate refractivity parameters at the different levels
Using Eq. (7), N can be replaced by NL, Nm, or Nu and at the regions using satellite data for the year
and the corresponding parameters ω and γ evaluated. 2007. This was first done annually (January –

Fig. 5 — Monthly mean variations of: (a) surface refractivity, Ns; (b) low level refractivity, NL; (c) mid-level refractivity, Nm; (d) upper-
level refractivity, Nu at the sub-Sahelian region
308 INDIAN J RADIO & SPACE PHYS, DECEMBER 2011

December); and then for the different seasons: dry is the actual distribution function. Using observed
season (November – March); and rainy season values, Fc (xj), j = 1,2,3,……. k, are the values of
(April – October). The results of these, as compared Fc(x) evaluated at x1≤ x2 ≤ x3 ≤…..≤ xk. It has been
with the actual and calculated values, are shown in chosen to compare the Dk values with the critical
Table 2. The agreement between them is remarkable. values of Dk for the 5 percent significant level
(ref 14,28).
3.3 Kolmogorov – Smirnov (KS) test
In order to determine whether any agreement exists 1.3581
D k (α =0.05) = …(9)
between the actual distribution function Fk(x) and k1/2
each generated series Fc(x), the Kolmogorov-Smirnov
(KS) test was applied (refs 13,22,24,28). The KS test Hence, if the value of Dk obtained exceeds that of
is used to determine whether agreement exists Dk(α = 0.05), the two functions are not close enough to
between the actual and calculated values. The be related. This criterion is used to judge the degree of
maximum deviation between Fk (x) and Fc (x) is: fit or relationship, if any, of the two functions under
consideration. On applying this test to the refractivity
Dk = max Fk (x) – Fc (x) …(8) parameters at different levels and for regions, the
results obtained are shown in Tables 3 and 4.
where, x, is ordered in ascending order such that 0 ≤ Hence, as shown in Table 3, the actual and model
x1 ≤ x2 ≤ x3 ≤……..≤ xk. functions are related at all the regions at the lower and
upper levels. At the middle level and for coastal and
For k observations, guinea savanna regions, their values showed
insignificance. This may be because, as is common to
Number of values ≤ x all surface based models, perturbations due to
Fk ( x) =
k disturbances, which originate (ref. 22) from within the
Table 2 — Application of proposed model for each of the regions annually and seasonally

(a) Coastal region k Ns NL=3.286+0.868Ns Nm=173.7+0.126Ns Nu=124.8-0.040Ns

(N-Units)
Actual Calculated Actual Calculated Actual Calculated
Annually 12 372.6 326.2 326.6 220.1 220.6 110.0 109.9
Dry season 5 369.7 324.1 324.1 215.8 220.3 109.8 110.0
Rainy season 7 374.6 328.4 328.4 223.3 220.9 110.1 109.8

(b) Guinea Savanna k Ns NL=3.231+0.868Ns Nm=77.21+0.382Ns Nu=109.5+0.001Ns

region
Actual Calculated Actual Calculated Actual Calculated
Annually 12 371.8 325.9 325.9 219.6 219.5 109.9 109.8
Dry season 5 382.0 334.8 334.8 221.8 223.4 110.0 109.8
Rainy season 7 364.5 319.6 319.6 218.0 216.7 109.9 109.8

(c) Midland region k Ns NL=89.02+0.69Ns Nm=149.8+0.209Ns Nu=108.5+0.004Ns

Actual Calculated Actual Calculated Actual Calculated
Annually 12 329.1 315.2 315.3 217.1 218.5 109.9 109.8
Dry season 5 293.7 290.9 290.9 209.8 211.2 109.8 109.7
Rainy season 7 354.3 332.6 332.6 222.3 223.8 109.9 109.9

(d) sub - Sahelian k Ns NL=95.73+0.691Ns Nm=161.7+0.173Ns Nu=109.6+0.009Ns

region
Actual Calculated Actual Calculated Actual Calculated
Annually 12 293.9 298.8 298.8 212.6 212.7 109.9 109.8
Dry season 5 250.2 268.4 268.6 205.3 205.1 110.0 109.8
Rainy season 7 325.2 320.5 320.3 217.9 218.1 109.8 109.8
 ADEYEMI & EMMANUEL: MONITORING TROPOSPHERIC RADIO REFRACTIVITY OVER NIGERIA 309

middle troposphere and whose effects are not quickly south movement of the intertropical discontinuity
communicated to the surface, may not be effectively (ITD). This is in consonant with the findings of
accounted for by the model. In addition, over the Balogun30 and Garbutt et al.31, who have used
hinterland of West Africa, the ITD surface is known radiosonde data obtained over Nigeria to investigate
to slope southwards with the moisture-laden south precipitation and precipitable water variations and
westerlies forming a wedge under a drier north- have concluded that differences in precipitation
easterlies29. Therefore, a station’s position relative to climatologies of the different regions in Nigeria
ITD is an important factor determining the structure account for the observed differences in the variations
of the moisture profile over the station. When a of their weather parameters; and that the variations
shallow moist layer is overlain by a deep dry layer, a observed in this weather parameters are dependent
model based on surface parameter may overestimate upon the north-south transect of the ITD. Applying
refractivity parameters aloft, and on the other hand, the analysis of variance (ANOVA) technique, a linear
where a dry layer is interspersed by a deep moist relation of the form
layer, an underestimation may result. These N = ω + γN S ( ω , γ , are constants)
limitations may then affect to a greater extent, the connecting the level refractivity parameters, N-units
effectiveness of the model. The actual and model (low-level refractivity, NL; mid-level refractivity, Nm;
functions for the mid-level refractivity during the dry and upper-level refractivity, Nu) with surface
and rainy seasons showed the same trend with it being refractivity, NS, have also been established from the
extended to the midland region (Table 4). satellite data for each region in Nigeria. The
difference observed in precipitation climatologies of
4 Conclusions the different regions has made it impossible to obtain
Using CM-SAF water vapour and temperature a single model that fits adequately the entire Nigeria
products retrieved from Advanced TIROS region. These relations when used to evaluate
Operational Vertical Sounder (ATOVS) onboard refractivity aloft obtained over each of the regions
National Oceanic and Atmospheric Administration during the year 2007 yielded an encouraging result.
(NOAA) satellites NOAA-15, 16, 18, it was possible On applying the KS test to the evaluated refractivity
to establish that seasonal variations of refractivity parameters, the actual and model functions are related
parameters over Nigeria is dependent upon the north- at almost all the levels considered at all the regions.
Table 3 — Result of KS test on actual and model results on
an annual basis Acknowledgements
The authors wish to express their gratitude to the
Region Dk (α=0.05) NL Nm Nu officials and management of the Department of
DL Dm Du Climate Monitoring SAF, Deutscher Wetterdienst,
Coastal 0.392 0.001 0.662* 0.130 Offenbach, Germany for providing the data. The
Guinea savanna 0.392 0.002 0.643* 0.034 efforts of Dr Jorg Schulz and Mr Markus Jonas of the
Midland 0.392 0.027 0.325 0.043 same Department, in making sure that the data were
Sub - Sahelian 0.392 0.069 0.292 0.048 available in ASCI format during the visit of the
*Association does not exist between actual and model values at corresponding author to CM-SAF is also greatly
5 percent significant level appreciated.
Table 4 — Result of KS test on actual and model NL, Nm and Nu values during the dry and the rainy seasons

Region Dk (α=0.05) Value of Dk for various refractivity Dk (α=0.05) Values of Dk for various refractivity
parameters during the dry season parameters during the rainy season
Dk for NL Dk for Nm Dk for Nu Dk for NL Dk for Nm Dk for Nu

Coastal 0.607 0.002 1.589* 0.311 0.513 0.000 0.557* 0.083

Guinea savanna 0.607 0.002 0.809* 0.072 0.513 0.004 1.103* 0.058
Midland 0.604 0.066 0.636* 0.069 0.513 0.034 0.558* 0.073
Sub - Sahelian 0.607 0.165 0.421 0.091 0.513 0.079 0.500 0.082

*Association does not exist between actual and model values at 5 percent significant level
310 INDIAN J RADIO & SPACE PHYS, DECEMBER 2011

References 17 Stergios A I & Thomas D X, Ten years analysis of

1 Kulshresta S M & Chatterjee K, Radio climatology of India: tropospheric refractivity variations, Ann Geophys (Germany),
Radio refractive index near the ground surface, Indian J 47 (4) (2004) pp 1333-1338
Meteorol Geophys, 17 (3) (1965) pp 367-384. 18 Fairall C W, A top down and bottom-up diffusion model of
2
2 Mitra A P, Reddy B M, Radicella S M, Oyinloye J O & CT2 and CQ in the entraining convective boundary layer, J
Feng S, Handbook on Radio propagation for tropical and Atmos Sci (USA), 44 (1987) pp 1009-1017.
sub-tropical countries (URSI Committee on developing 19 Olaniran O J & Sumner G N, A study of climatic variability
countries UNESCO subvention, ICTP, Italy), 1987. in Nigeria based on the onset, retreat, and length of the rainy
3 Willoughby A A, Adimula I A & Sorunke A O, season, Int J Climatol (UK), 9 (1989) pp 253-269.
Characteristics of surface radio refractivity over Ilorin, 20 Aro T O, Analysis of data on surface tropospheric water
Nigeria, Nigeria J Phys, 15 (1) (2003) pp 45-50. vapour, J Atmos Terr Phys (UK), 38 (1975) pp 565-571.
4 Crane R K, Fundamental limitations caused by RF 21 Adedokun J A, Intra layer/mid tropospheric precipitable
propagation, Proc IEEE (USA), 69 (2) (1981) pp 196-209. water vapour relations and precipitation in West Africa, Arch
5 Anderson K D, Interference of refractivity profiles by Metorol Geophys Bioclimatol B (Austria), 33 (1983)
satellites ground RF measurements, Radio Sci (USA), 17 (3) pp 117-130.
(1982) pp 653-663. 22 Adedokun J A, On a relationship for estimating precipitable
6 Anthony R C, Chris Rocken, Sergey V, Solokors Kij & water vapour aloft from surface humidity over West-Africa,
Kenneth D Anderson, Vertical profiling of atmospheric J Clim (USA), 6 (1986) pp 161-172.
refractivity from ground based GPS (American Geophysical 23 International Telecommunication Union-
Union, USA), 2002. Radiocommunication sector (ITU-R), The radio refractivity
7 Adeyemi B, Surface water vapour density and tropospheric index: its formula and refractivity data, Rec. ITU-R P453
radio refractivity linkage over three stations in Nigeria, J (ITU, Geneva), 1997.
Atmos Sol-Terr Phys (UK), 68 (2006) pp 1105-1115.
24 Adedokun J A, Surface humidity and precipitable water
8 Bean B R & Dutton F J, Radio meteorology, National Bureau
vapour linkage over west and central Africa: Further
of Standard monographs 92 (US Department of Commerce, clarification and evaluation of existing models, Int J Clim
Washington), 1966. (UK), 9 (1989) pp 425-433.
9 Hall M P M, Barclay L W & Hewit M T, Propagation of
25 Agunlejika & Raji, Electrical evaluation of water term of
radio wave (IEEE, USA), 1996.
refractivity in Nigeria, Int J Eng Appl Sci (Nigeria), 2 (2)
10 Saha Korak, Parameswaran K & Raju C Suresh,
(2010) pp 63-68.
Tropospheric delay in microwave propagation for tropical
26 Adedokun J A, West African Precipitation and dominant
atmosphere based on data from the Indian subcontinent, J
atmospheric mechanisms, Arch. Meteorol Geophys
Atmos Sol-Terr Phys (UK), 69 (2007) pp 875-905.
Bioclimatol A Meteorol Geophys (Austria), 27 ((1978) pp
11 Adediji A T & Ajewole M D, Vertical profile of radio
289-310.
refractivity gradual in Akure Southwest Nigeria, Prog
27 Adeyemi B & Babalola Y A, Diurnal variations of absolute
Electromagn Res (Hong Kong), 14 (2008) pp 157-168.
humidity at Ibadan and Minna, Nigeria J Sci Eng Technol, 7
12 Babalola M T, Studies on the vertical model of of the radio
(2) (2000) pp 2371-2375.
refractivity in Nigeria, Africa J Sci (Nigeria), 2 (1) (1998)
28 Kendall M & Stuart A, The advanced theory of statistics Vol
pp 1-10.
2, 4th edition (Charles Griffin, London), 1979, pp 476.
13 Adeyemi B, Tropospheric radio refractivity over three
radiosonde stations in Nigeria, Ife J Sci (Nigeria), 6 (2) 29 Leroux M, La dynamique des Precipitations en Afrique
(2004) pp 167-176. occidental Asecna 39 Dakar Senegal, J Clim (USA), 6 (1971)
pp 161-172.
14 Adeyemi B & Adedayo K D, Atmospheric radio refractivity
and water vapour density at Oshodi and Kano, Nigeria, 30 Balogun E E, Seasonal and spatial variations in thunderstorm
Nigeria J Pure Appl Phys, 4 (1) (2005) pp 11-12. activity over Nigeria, Mon Weather Rev (USA), 36 (7) (1981)
15 Bean B R & Thayer G D, Models of the atmospheric pp 192-197.
refractive index, Proc Inst Radio Eng (USA), 47 (5) (1959) 31 Garbutt O J, Stern R D, Dnnett M D & Elston J, A
pp 740-755. comparison of the rainfall climate of eleven places in West
16 Bayong T H & Djakawinata S, The influenced of Africa using a two-part model for daily rainfall, Arch
meteorological factor, on tropospheric refractive index over Meterol Geophys Bioclimatol B (Austria), 29 (1981)
Indonesia, J Mater Sci (USA), 4 (1) (1999) pp 1-12. pp 137-140.