The Challenges to Nigerian Raingauge Network Improvement

This research is geared towards establishing the link between the need for Nigeria rain gauge network improvement and the low efficiency of the current operating system with their various challenges. Nigeria currently has 87 existing and operating rain gauges while it need extra 970 gauges for proper rainfall measurement and achievement of gauge density of 874 km 2 per gauge. According to the World Meteorological Organization guideline, the required number of rain gauges to minimize cost and maximize efficiency for temperate Mediterranean and tropical flat area is in the range of 600-900 km 2 per gauge. The impact of not increasing Nigeria existing rain gauges is a 10% increase in error of any design based on the current 87 gauges. The flood devastation of the country in September 2012 due to high rainfall and release of flood water from dams are examples of effect of improper planning arising from no or inaccurate records. Equally of great importance is the negative contribution of such error to our economic development as a nation with failure of existing infrastructure being recorded in our roads, rail lines, air ports, water ways and other water resources infrastructures. Finally, it is noteworthy to observe that these largely man-made problems can be overcome with leadership that is alive to its obligation to the society it governs as we face the challenges of globalization and global climate change.


INTRODUCTION
The Nigerian Meteorological Agency is by law charged with the responsibility of establishing the rain gauge network as part of the country's water and land information systems.The national data set formed is used to support climate mapping, flood early warning and other hydrological studies, while it is expected that the information is freely accessible to any interested party.The objectives of this study are summarized as follows: • Assessing the existing rain gauge network in Nigeria for adequacy • Studying the challenges of network density improvement It is important to realize that measurements made in a rain gauge represent samples over an area.For this reason, the rain gauge must be located and installed so that its catch is as representative as possible of the area being sampled.Particularly, determination of mean rainfall falling on a watershed is fundamental to many hydrologic studies.It must be determined as accurately as possible, but its accuracy is highly dependent on the rain gauge network design.
An ideal method developed for designing the rain gauge network would not only allow improvement of the accuracy of mean rainfall estimation but would also have flexibility useful for relocating the gauges and improving the gauge density according to the expected level of statistical accuracy.
Almost every aspect of Nigeria development is closely dependent on an adequate supply of water (Sonuga, 1981;Revenga and Cassar, 2002).The draught of the nineteen seventies that swept most of the Sahel region of Africa, including Nigeria, has left the region with general water resources scarcity (Offordile, 1992;Batterbury, 1998).The consequence of this is low Agricultural out-put, limited water supply and inadequate water reserve for power generation that the region is faced with.This means that depletion of the few resources available would be imminent if appropriate management measures are not put in place (Tanko, 1996).For proper management of the flow water resources in Nigeria, the then Federal Military government established River Basin Development Authorities in 1976.These River Basins were charged with major objectives of water resources development and planning for more agricultural yield, abundant water for domestic supply, power generation, communication, flood/erosion control and so on.These objectives are stipulated in below section, subsection (1) of Decree No. 25 of 1976 which was modified in section 4 of Decree No. 87 of September 1979 and Act No. 7 of October 1981.To enable the attainment of the  (Hendrick and Comer, 1970).Most River Basins in Nigeria and the Nigeria Meteorological Agency estimate their rainfall information using a network of manually operated rain gauges.These rain gauges are located mostly in city Centre's and project location thus, information on rainfall for water resources planning and studies which need high precision, may not be achieved.This is justified by the fact that the present network operates at a density lower than the minimum density established by the world metrological organization (WMO, 1965) for tropical zones, Mediterranean and temperate zones of 600-900 km 2 /gauge.The density with the highest number of gauges (65 in 1987) in the Upper Benue River Basin, that is, 1906 km 2 /gauge.The reality on the ground today indicates that the values are much lower at 17 existing and operation stations, as observation stations are being closed down due to paucity of fund for installation and maintenance of both old and new stations to the detriment of accuracy in estimation.Therefore, in network design problem, accuracy in estimation and cost are the two controlling factors (Ngene, 2009).
As reported by Ologunorisa (2009) very few studies have been carried out on content analysis of hydro meteorological networks in the tropics.For the few studies, it is observed that location of stations was based more on political reasons than the need for appropriateness in achieving stated objectives.(Oyebande, 1990) using the WMO Standard gave the optimum number of rain gauges for Nigeria as 1074.For this study, the country was divided into sections with geological and hydro meteorological similarities.The (Yemag, 1992) project was however for the upper Benue River Basin alone with an area/gauge ratio of 900-3000 km 2 /gauge.(Tanko, 1996) used Shihs' principle to estimate the area/gauge ratio of 2510 km 2 /gauge for the same River Basin which is still above the WMO Standard.
The objective of this work is to situate the inadequate number of rain-gauge stations in Nigeria to the various challenges of establishing and running the stations.

MATERIALS AND METHODS
Precipitation: Precipitation is the primary input vector of the hydrologic cycle.Precipitation includes all forms of moisture falling from the atmosphere to the earth's surface, primarily produced from water vapour present in the air.
In tropical countries such as Nigeria, rain forms the bulk of precipitation as such the word "precipitation" can be interchangeably used with rainfall (Tanko, 1996).
There is a wide variety of rain gauge types.However, to select a rain gauge for a particular watershed, the best criterion, which mostly governs the choice, is compatibility with other gauges being used within a country.
In Nigeria, the installation and observation of rain gauges throughout the country are controlled by the department of meteorological services, Federal Ministry of Aviation with its headquarters in Oshodi, Lagos-Nigeria.The department has been collecting, archiving, interpreting weather and preparing climate information for many years.Three gauges used for the measurement of rainfall are: • Non recording gauges • Recording gauges • Weather radars Errors of rainfall measurement: Rodda (1967) noted that despite early progress in rainfall measurement and advancement in technology, rainfall measurement results is still marginally near to the solution or the true value.The difference between the measured and true rainfall is attributed to measurement errors.
These are broadly classified by Ward (1975) and Morley and Ismail (1981) as systematic error and random errors.Systematic errors, they observed, are  , 1974).

Rain gauge network density:
A precipitation gauge measures the rainfall at one geographical point and cannot be representative of the precipitation on a larger area except in its immediate vicinity.The larger the area, great the error in the assumption because meteorological conditions may occasionally produce intensities at a point greater than any possible combination of circumstance could produce over an area greater than a few hectares (Mutreja, 1990).Rain gauges are used in measuring the amount of water that falls over given watershed or catchments.Using a number of well-planned rain gauges over a catchment constitute the network.The catchments area in kilometer square per the number of gauges gives the network density.Arising from too dense a network is the issue of cost implication or the high error resulting from low-density network.
In his research work, (Ngene, 2009) used Shih approach to design the precipitation network for Nigeria, Table 2. Accordingly, the 87 existing operating gauge stations need to be upgraded with 970 more gauges for proper rainfall measurement.From this study, it is also known that 204 gauges stations are existing but need to be reactivated by equipping them with new rain gauge instrument.This implies that only 766 new stations need to be established in the country to meet with world meteorological organization requirement for temperate Mediterranean and tropical flat area.
Data collection: Information for network improvement was obtained from the Nigeria Meteorological Agency, various River Basin Development Authorities and some State Agricultural Development projects.The data consist of monthly rainfall records for existing network, with ten years consistent records considered (Fig. 1).
The study area is Nigeria, which is located within latitude 4°-14°N and longitude 2°-15°E with an area of 923,378 km 2 (MOW, 2014).The climate within this area is relatively uniform with only the major Rivers providing important differentiation of vegetation type and agricultural practice (Gbuyiro et al., 2002).The climate of Nigeria can be conveniently classified into Rainy and Dry seasons.Each of these seasons last for six months on the average i.e., April to October for rainy seasons and dry season is November to March.The main annual rainfall ranges from less than 500 mm in the extreme north to about 2500 mm along the coast.

Theory of analysis:
For each catchment area using the River Basins, the area is subdivided into n strata.The mean rainfall for each station is: where, = The time interval, R ki = Mean rainfall of the K th station in the ith stratum M ki = The total number of observation ˞ = The recorded rainfall The mean rainfall for a whole zone can be given by: where, a i is the weighting ratio or ratio of stratum area to the area of the catchment.From these data, the variance and covariance is also obtained.The precision reliability standard error from a two stage, statistical sample theory was given by Shih (1982) where, τα values are taken from a t-table for α confidence level with (˚− 1) degree of freedom; β the For stratified sample method with optimum allocation, combining relative variance with standard error formulae and making; β and ˚ the subject matter: And: ( ) where, α and β are the desired levels and statistical accuracy chosen based on factors such as impacts of land use conditions, population density, economic development and the importance of water resources in the country (Shih, 1982).Raghunath (1986) reported a simple approach that relates the number of gauges to the coefficient of variation and error to be tolerated.The relation was of the form: This approach though simple is insensitive to the distribution of the gauges and the catchment area.The two approaches as described above were used in designing the Nigerian Rain gauge network density.

DISCUSSION OF RESULTS AND CHALLENGES DISCUSSION OF RESULTS
From optimization results for Anambra/Imo river basin shown in Table 3 which also apply for other basins, stratified with optimum allocation of gauges estimate mean rainfall better than random sampling method and stratified without optimum allocation of gauges.Spatial variance for random sample and stratified with and without optimum show closeness.This closeness is based on the assumption that spatial variance is a hydrological characteristic, that is, independent of the sampling method.Stratified method with optimum allocation of gauge gives lower percent relative error at low mean rainfall than the other two methods.The current study have shown that at reduced  • Proper design of a network of rain gauges in a given watershed to collect the necessary precipitation.• Though rainfall information could be estimated using weather sensors such as satellites and radars, cost of installation, operations and maintenance especially in developing country, such as Nigeria, is high.• All the River Basins in Nigeria, State Agricultural projects, schools and colleges and indeed Nigeria Meteorological Centre depend on rain gauge network for rainfall information estimation which accuracy depends on the density used and locations.• It is important to continuously check the precision of the existing network in estimating rainfall and establish an optimum network.• Due to lack of complete rainfall data for some stations, interpolations are used to cover up for missing ones.• This study has shown that as a result of the low number of existing and operating rain gauges.The rainfall information collected in Nigeria may not be reliable.• Part of the maintenance difficulty faced at Tiga dam in Kano after its construction in 1978 was the inadequacies of the original design arising from poor hydrological investigation and the flooding of nearby Bagauda earth embankment dam (Eadie et al., 1992).• Again the flood disaster that has claimed life and properties around Ogunpa River in Oyo station in 1980 may be attributable to inadequate rain gauge (Nigeria Tribune 13 th August 2011).This is to enhance prediction of the volume of expected flood and hence a design to properly channel the floodwater.

LIMITATIONS TO DATA COLLECTION IN NIGERIA
Quite a lot of obstacles are laid on the way of any individual seeking data for any purpose in this country even from government agencies, individuals and corporate world.Primary data is very important in any research endeavor as it form the basis of analysis.Wrong or inaccurate data therefore increases the chance of arriving at wrong result or hypothesis.The following list is an indication of the areas an individual sourcing for data in Nigeria must not overlook when sourcing for information for any research work.

Logistics:
• High cost of traveling between cities for data that can be located in one office within an organization.• Vast nature of the country hence visiting many cities for data that can be centrally located.• Haphazard address system sometimes lead to delay in finding the location of offices and sites.

Data:
• Long term records are sometimes lost in transit, hence short term records with potential errors are used for analysis.• Little or no electronically digitalized storage, retrieval and use of data in government establishments.• Data are sent to Nigeria Meteorological Services in Lagos with no copy kept at regional offices.

Data processing and storage:
• Movement of data materials from one place to another without proper records and referencing.• Some gauges are operated for the meteorological agencies by individuals who are poorly paid without any attachment, hence no commitment to the project and are possible error sources.

Data retrieval:
• No central pool where data could be sourced in a global village of the 21 st century.
• Haphazard manual record keeping resulting in loss of valuable time and energy in locating data where available.• Low internet use by individual, corporate and government agencies.Rainfall data should be uploaded at agencies website for easy of retrieval though at cost to the user.

Human factor:
• High endemic corruption in the system which affects all levels of government activities.• Perception of information seekers as pest who should at best be drained of all the money they can afford to give.• Jealousy, resulting to researchers being seen as possible competitors.• Nonexistent patriotic spirit to help researchers for the overall good of the nation through their findings.
• Absence of schedule officers sometimes mean making the journey of a researcher a second visit inevitable as other persons may not be in a position to give out required information.• General lack of good work attitude as it concerns handling of data.

IMPACT OF OPTIMIZATION OF NIGERIA RAINGUAGE NETWORK
As earlier pointed out, Nigeria has only 87 existing and operating gauge stations out of the required 1057 stations.This leaves the country with a balance of 970 stations to be provided, inclusive of the 204 station that exist but not operating.
Not having adequate number of station is indicative of the reasons for the failure of structures such as dams, river bridges, poor agricultural yield and little or nonexistent flood estimation and control measures.The flood disaster that ravaged 9 States in Nigeria in September 2012 is an example of failure of planning due to lack of adequate rainfall data.Using (WMO, 1965) empirical formula, the Fig. 2 shows the error associated with Nigeria current number of 87 existing gauge stations and the desired 1057 gauge.
From the graph above, the error of estimate associated with the current network that is 87 rain gauges will give a 14.1% risk which can impact heavily on designs based on this, on say, culvert carrying capacity, dams and the release of water downstream, railway lines and highway and bridges.However, when the number of gauges increases to say 970 stations, the error stabilizes at around 4%.
Example and result: If a flood over a culvert is designed with a risk of 4% and the design life is 15 years, the return period of the flood on the culvert is estimated at: where, ˞ = 4% and J = 15 And T = 1/p = 368 years However, if the risk, R is increased to 14% with the same design life of 15 years, the return period of the flood on the culvert will reduce to 100 years.This indicates 4 times probability of occurrence hence more likelihood of failure within the design life of the structure.

CONCLUSION AND RECOMMENDATIONS
A new method developed by Shih was used to determine the optimized number of rain gauge station required for Nigeria to be 1057 stations with the gauge density of 874 km 2 /gauge.This value falls within the world meteorological organization minimum density range of 600-900 km 2 /gauge for temperate Mediterranean and tropical zone flat area.To further improve on the density or strength on data collection for rainfall information the following recommendations are made: • All twelve River Basins Development Authority should serve as the zonal headquarters, while the Nigerian Meteorological agency Lagos office will be the final authority on weather matter, that is, data collection, storage and retrieval.

Fig. 1 :
Fig. 1: Rain gauge stations in Nigeria degree of accuracy.For stratified sample method with optimum allocation, combining relative variance with standard error formulae and making; β and ˚ the subject matter:

Table 1 :
River basin development authorities in Nigeria and their headquarters locations

Table 2 :
Number of existing and operating stations against watershed area

Table 3 :
Comparisons of regression estimates for the relative standard error of monthly mean rainfall for Anambra/Imo River basin • Other state offices including the ministries and their agencies such as ministry of Agriculture, research institutions, schools and colleges should route their reports through the various River Basin collection centers in the state.•The current Nigeria Meteorological Agency should be adequately equipped and commercialized.Education and training of staff are therefore advocated.The equipment needs regular inspection and calibration to ensure that they are always at optimal use.• Satellite and radar coverage of the country is advocated to compliment the rain gauge arrangement.• Budgetary provision for procurement and installation of new rain gauge station should be made and their effective utilization ensured.• The cost of opening 970 new rain gauge stations which include the existing but not operating stations is put at N92 million.