Assessment of the Skill of the Cmorph Technique in the
simulation of Rainfall over Kenya.
Okuku C C1 Rarieya O P2 Mutemi J N3
1 Kenya Meteorological department P O Box 30259-00100 Nairobi Kenya
2 IGAD Climate prediction and Application Center P O Box 10304-00100 Nairobi Kenya
3 University of Nairobi, department of Meteorology P O Box 10304-00100 Nairobi Kenya
ABSTRACT
In this study verification of cmorph rainfall data and the observed were examined. The study was
carried out on seasonal time-scale. Graphical comparison of the cmorph rainfall data and the
ground observation was used in order to achieve the objectives of the study. The data sets
included observed daily rainfall from rain gauge and the daily cmorph rainfall data from
satellite. The period of study was from October to November 2005. During the study cmorph
data and the observed rainfall data produced a similar graphical presentation showing that the
cmorph technique can be used to extract daily rainfall over Kenya. Cmorph rainfall data over
coastal area showed a fine graphical presentation with the observed rainfall for some selected
stations over the coastal region. Hence it was concluded that there was a strong similarity
between the observed rainfall and the cmorph data from the satellite.
INTRODUCTION
Kenya is located on longitudes 34°-42° E
and latitude 3.5°N/S. It experiences two rain
seasons i.e. March, April and May long rain
season and September, October and
November, short rain season. Her economy is
mainly dependent on agriculture and agro based
industries which heavily rely on the amount and
distribution of weather elements such as rainfall,
temperature and humidity. Accurate analysis
and monitoring of these weather/climate
elements is then used in prediction and early
warning of the climate extremes to improve
planning and management of climate sensitive
activities and reduce the associated socio
economic miseries that are prevalent in the
region. There is sparse ‘instu’ rainfall
measurement instruments in developing
countries; inadequate data for use in analysis and
prediction. See map below.
Map of Kenya showing synoptic stations
20
15
october
10
november
5
31
29
0
17

Daily rainfall data over Kenya for
October and November for the year
2005 was obtained from Kenya
Meteorological department(KMD)
Daily Cmorph data for October and
November for the year 2005 was
obtained
from
NOAA
CPC
morphing technique (CMORPH).
Daily rainfall during October-November
2005
Most parts of Kenya experienced rainfall
during this season. Most stations recorded
rainfall less than 50% of their respective
October –November total rainfall. The only
stations which recorded normal to above
normal were Mombasa, Makindu and
Machackos.
The figures below show the daily rainfall of
some chosen stations. The stations were
chosen on the basis of having received
widespread rainfall.
27

Results and Discussions
25
Data type and source
refer to this latter step as "morphing" of the features.
23
It should be noted with a lot of concern that
extreme climate events affect the welfare of
communities and tend to enhance poverty
especially the rain-fed agricultural and
hydroelectric power firms, major sources of
food and energy respectively. These entire
extreme Weather/climate events can be
reduced through good understanding of past
climate, enhanced monitoring and timely
early warning as well as awareness of the
usefulness of climate information and
prediction products in decision making. This
can only be achieved by using a wellanalyzed captured data. In this study we
have compared the accuracy of cmorph
rainfall data and the observed ones on the
ground to create confidence for the users of
this rainfall data.
Propagation vector matrices are produced by
computing spatial lag correlations on successive
images of geostationary satellite IR which are
then used to propagate the microwave derived
precipitation estimates. This process governs the
movement of the precipitation features only. At
a given location, the shape and intensity of the
precipitation features in the intervening half
hour periods between microwave scans are
determined by performing a time-weighting
interpolation
between
microwave-derived
features that have been propagated forward in
time from the previous microwave observation
and those that have been propagated backward
in time from the following microwave scan. We
21
Justification for the study
Model description
19
The main objective of this study is to verify
the cmorph rainfall data and the rainfall
observations made on the ground.
Specific objectives include:
1. Assessing the accuracy of the
cmorph technique comparing its
rainfall output with the observed
rainfall at 24hr interval.
2. Determine the spatial variation of
rainfall for both the cmorph
technique and the observations.
 Model description,
 Data generation using GRADS
 Graphical analysis
The verification period was from October
17th to November 25th the year 2005
rainfall
Objectives of the study
days
Methodology
The methods used included
Figure 1.0 daily rainfall for Makindu
25
raifall
20
15
october
10
november
5
31
29
27
25
23
21
19
17
0
days
Figure 1.1 daily rainfall for Mombasa.
Figure 1.4
Figures 1.2 and 1.3, show rainfall
distribution on 23 and 24 November.
The figures below clearly show that there is
a harmonized relationship between cmorph
rainfall and the observed. Given that cmorph
technique best gives spatial distribution of
rainfall of a given place at some period, the
peeks below show that there was high
amount of rainfall at different regions during
the rainfall occurrence.
coastal cmorph distribution
6
Figure 1.2
rainfall
5
4
3
2
1
Figure 1.2 shows the spatial variation of
rainfall on 24th November 2005. The
technique represents well cmorphed spatial
distribution of rain.
0
1
4
7
10 13 16 19 22 25 28 31 34 37 40
Figure 1.5 Lamu Cmorph 2005
Lamu
rainfall
15
10
Series1
5
0
1
5
9 13 17 21 25 29 33 37 41
Days
Figure 1.6 Lamu Observed
coastal cmorph distribution
Figure 1.3
6
Figure 1.3 shows the spatial variation of
rainfall on 23 November. Similarly during
the month of October some days recorded
more rainfall as shown below. Cmorph
precipitation for 21 October 2005.
rainfall
5
4
3
2
1
0
1
4
7
10 13 16 19 22 25 28 31 34 37 40
Figure 1.7 Voi Cmorph
3
Voi
rainfall
15
10
Series1
5
0
1
5
9 13 17 21 25 29 33 37 41 45
days
Figure 1.8 Voi Observed
Cmorph lon 34 - 38 lat 2S - 2N
6
rainfall
5
4
3
2
1
0
1
3
5
7
9
11
13
15
17
Figure 1.9 Dagoretti C-MORPH
Daoretti
rainfall
60
40
Series1
20
0
1
5
9 13 17 21 25 29 33 37 41 45
days
Figure 2.0 Dagoretti Observed
Cmorph lon 34 - 38 lat 2S - 2N
6
5
rainfall
cmorph technique was well captured. The
techniques over and below estimation of
rainfall was also observed due to the proxy
data obtained from satellite sensors. Cmorph
rainfall data over coastal area showed a fine
graphical presentation with the observed
rainfall for some selected stations over the
coast. Hence it was concluded that there was
a strong similarity between the observed
rainfall and the cmorph data from the
satellite creating confidence in the
technique.
4
3
2
Recommendations and suggestions for
further studies.
The present study of the cmorph technique
shows that the technique is promising.
However it has some weaknesses of
overestimation and underestimation of
precipitation. It is therefore recommended
that improvement on the sensors temporal
and spatial resolution is checked. With
regard to spatial resolution, the precipitation
estimates are available on a grid with a
spacing of 8 km (at the equator), the
resolution of the individual satellite-derived
estimates is coarser than that - more on the
order 0f 12 x 15 km or so. Hence it is
recommended that a technique with a much
finer resolution is used in order to obtain the
best precipitation data.
1
0
1
3
5
7
9
11
13
15
17
Acknowledgements
Figure 2.1 Makindu C-MORPH
rainfall
Makindu
20
15
10
5
0
1
4
This work was carried out with funding
support of WMO and presented to CIMO
Russia 2008. The authors are also grateful to
Kenya government (KMD) and the
University of Nairobi for providing
necessary resources.
7 10 13 16 19 22 25 28 31 34 37 40 43
days
Figure 2.2 Makindu Observed
References
Conclusion
October-November season had a depressed
rainfall due to diffuse ITCZ over the region
of study. The spatial distribution of the
Ferraro, R. R., 1997: SSM/I derived global
rainfall
estimates
for climatological
applications. J. Geophys. Res., 102, 1671516735.
Ferraro, R. R., F. Weng, N. C. Grody and
4
L. Zhao, 2000: Precipitation characteristics
over land from the NOAA-15 AMSU
sensor. Geophys. Res. Ltr., 27, 2669-2672.
Joyce, R. J., J. E. Janowiak, P. A. Arkin,
and P. Xie, 2004: CMORPH: A method that
produces global precipitation estimates from
passive microwave and infrared data at high
spatial and temporal resolution.. J.
Hydromet.,
5,
487-503.
Kummerow, C., Y. Hong, W. S. Olson, S.
Yang, R. F. Adler, J. McCollum, R. Ferraro,
G. Petty, D-B Shin, and T. T. Wilheit, 2001:
Evolution of the Goddard profiling
algorithm (GPROF) for rainfall estimatin
from passive microwave sensors. J. Appl.
Meteor., 40, 1801-1820
5