Research Article | | Peer-Reviewed

Performance Comparison of Empirical Models for Estimating Global Solar Irradiation in the Soudano-Sahelian Zone of Cameroon: The Case of the City of Maroua and Garoua

Received: 8 June 2024     Accepted: 1 July 2024     Published: 15 July 2024
Views:       Downloads:
Abstract

The main objective of this study is to compare thirty-five (35) solar radiation models available in the open literature in order to predict monthly solar radiation in two main cities of Cameroon. This estimation and comparison are based on selected statistical comparison parameters named, root mean square error (RMSE), mean bias error (MBE), mean percentage error (MPE) and determination coefficient (R2). These different models are implemented using regression analysis tools named Exel and MATLAB. Estimated values were compared with measured values according to normalized values of statistical parameters, using measured meteorological data of more than 19 years, from 1984 to 2015. All the models have been classified with their associated ranking according to their statistical parameter accuracy. From this study it appears that the models of Ertekin and Yaldiz (MOD20), Togrul and Onat (MOD28), are more accurate than other models. Indeed, for the city of Maroua (MBE%=-2.82E-14; RMSE%=0.862; MPE=-0.00845; R2=0.985), while for Garoua (MBE%=-9.21E-15; RMSE%=0.806; MPE=-0.00631; R2=0.959). according to their accuracy these models can be therefore be used to predict monthly solar radiation for soudano-sahelian regions of Cameroon. Correlation equations found in this paper will help solar energy researcher to estimate data with trust because of its fine agreement with the observed one. hence the models presented in this study could be used to evaluate accurately the solar radiation at any locations with similar climate.

Published in International Journal of Sustainable and Green Energy (Volume 13, Issue 2)
DOI 10.11648/j.ijrse.20241302.12
Page(s) 28-42
Creative Commons

This is an Open Access article, distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution and reproduction in any medium or format, provided the original work is properly cited.

Copyright

Copyright © The Author(s), 2024. Published by Science Publishing Group

Keywords

Meteorological Data, Global Solar Radiation, Empirical Models, Performance, Statistical Parameters, Cameroon

1. Introduction
As a result of the increasing number of solar energy applications, the need for solar radiation data became more and more important. There is a growing need for quality data to facilitate studies of solar systems around the world. Quality data are important to design and to optimize solar energy conversion systems. Solar data also helpful when evaluating the techno-economic feasibility of the project, thereby helping the investors, government agencies and the utility operators for an informed decision making . Unfortunately, in developing countries, researchers encounter difficulties with data gaps relative to the scarcity of data records of stations or the continuity of readings, even in many developed countries there is a dearth of measured long-term solar radiation data . The amount of global solar radiation at any site is best determined through the installation of measurement instruments, such as the pyranometer, for the monitoring and storing of its day-to-day recording, but it is a very costly and tedious . Thus, it becomes necessary to look up procedures in order to supply solar radiation data estimation for area with the gaps in the measurement records or where measurements are not carried out by using empirical models. Input variables used in empirical models of solar radiation generally include sunshine hours, mean temperature, maximum temperature, minimum temperature, number of rainy days, extraterrestrial radiation, cloudiness, soil temperature, altitude, latitude, relative humidity, albedo, precipitation, and evaporation . These related models are generally presented as linear, logarithmic, exponential and hybrid, quadratic, quartic, cubic, power forms .
Parameter used earlier to evaluate the overall solar radiation is the effective sunshine duration. This duration is easily measure in hours with Campbell stokes heliograph. the simplest model to estimate solar radiation using sunshine duration, is the model of Angstrom establish in 1924 , and their associated models establish by Prescott in 1940 known as Angstrom-Prescott-Page model . With the models mentioned and for different sites around the world many researchers found the regression coefficients within reasonable degree of accuracy . One drawback of the model of Angstrom-Prescott-Page is the fact that it took little input data while for many regions around the world evaluation of the solar radiation need more magnitudes than sunshine hours. To fill this gap many others models have been developed by researchers around the world. The models are classified depending on available meteorological data.
Despite the fact that numerous works relative to the development and improvement of empirical correlation for determination of monthly averaged daily global solar radiation in locations around the world, no more correlations are found for the regions of Cameroon, except Hargreaves and Samani model, angstrom-Prescott model, Bristow and Campbell model Annandale et al. Model, and Goodin et al. Model, , it is also found that for the city of Garoua and Maroua, the global solar radiation data have not been studied seriously. It is on the basis of these observations and depending on the available collected data, that we decided to select thirty-five (35) representative models amongst those encountered in open literature. This paper deal with the evaluation of the performance of the mentioned models for two different stations in Cameroon. To attain this goal, Excel and Matlab, tools were used to find out at first, the regression coefficient of the models and secondly the strength of statistical performance parameters named, root mean square error (RMSE), mean bias error (MBE), mean percentage error (MPE), and determination coefficient (R2), in order to know how performant selected models are.
1.1. Study Area and Data
The study area covers two administrative regions of Cameroun the climate is warm semi-arid (BSh) climatic types (according to Köppen-Geiger climate classification system) as presented in table 1. Meteorological data for this study were collected from these stations corresponding to each administrative region. Data and the recorded time period are shown in Table 2.
Table 1. Geographical coordinates of the different stations.

location

Climate Zone

Latitude(°N)

Longitude (°E)

Elevation (m)

Maroua

Warm semi-arid (BSh)

10°28’N

14°16’E

423

Garoua

Warm semi-arid (BSh)

9°20’N

13°23’E

241

Table 2. Available measured meteorological data associate with record time.

Meteorological data denomination

period

Record time (years)

Daily temperature (maximum)

1980 to 2013

21

Daily temperature (minimum)

1980 to 2013

21

Daily temperature (mean)

1980 to 2013

21

Soil Temperature (mean)

1980 to 2013

21

relative humidity (mean)

1980 to 2013

21

precipitation (mean)

1980 to 2013

21

Effective day length (mean)

1961 to 2015

33

Monthly Solar radiation (mean)

1984 to 2015

4

1.2. Sources of Solar Radiation Databases
There are three types of measurement of the solar radiation data, in practice none of them appear to be perfect. It is therefore significant to know the strengths and weaknesses of each type. The three main sources of data on solar radiation at the earth surface are: Ground measurements, satellite data-based calculations, and empirical models based on mathematical equations. These mathematical equations use meteorological data as input parameters (relative humidity, temperature, sunshine hours, soil temperature, altitude, number of rainy days, total precipitable water, albedo, latitude, cloudiness and evaporation) The advantages and limitations of these different data sources are presented in Table 3.
Table 3. Strengths and weaknesses of radiation databases .

Type of measurement

Advantages

Limitations

Ground measurements

1) High accuracy at the point of measurement

2) Good measurements frequency

3) High-quality data (in the rigorously controlled and managed conditions)

4) Redundant measurements enable more stringent quality control

1) Limited measurement time

2) Number of measurement sites limited

3) Unknown accuracy (in historical data)

4) Different periods of measurement

5) Maintenance Operation of a ground station (

6) Regular maintenance and calibration)

7) Management of data from many different providers

8) Representation is limited by geography and the level of data aggregation.

9) High costs for acquisition and operation

Satellite-derived data

1) Available everywhere (continuous global coverage)

2) Spatial resolution from 3 km

3) Frequency of measurements from 15 minutes

4) Spatial and temporal consistency

5) High calibration stability

1) Poor instantaneous accuracy for the point estimate (in comparison to high quality ground measurements)

2) Time step 15 and 30min

3) Representation of the area (typically a grid cell 3 to 6 km)

Mathematical Models

1) Available according to the model and data involve in the model

2) Models can be extended to other similar sites

3) No direct measurements needed

1) Low accuracy

2) Accuracy depending on the type of model and local climate

2. Materials and Methods
Depending on the meteorological data, thirty-five models are selected amongst those available in open literature for the regression analysis. Regression coefficients are then generated from regression analysis for each model. The normalized values of MBE, RMSE, MPE and determination coefficient (R2), are determined in order to know the performance of each model. This is made possible by using Excel and Matlab regression tools.
2.1. Studied Models
Global solar radiation from empirical models can be classified into four subgroups (temperature-based, cloud-based, sunshine based, and hybrid parameter-based models). Models selected usually takes into consideration two features: (1) available meteorological input data (2) the model accuracy. Selected models are presented in the “Table 4.”
Table 4. Thirty-five selected (35) empirical models, their equation and types of variables.

Mathematical equation

equation type

Authors and reference

MOD01

HH0=m+nSS0

Linear equation

Page 1961

MOD02

HH0=mcosφ+nSS0

Linear equation

Glower and McCulloch 1958

MOD03

HH0=m+nSS0+oSS02+pSS03

Cubic equation

Samuel 1991

MOD04

HH0=m+n*logSS0

Logarithmic equation

Ampratwum and Dorvlo 1999

MOD05

HH0=m+nSS0+oφ+pSS0

Linear equation

Dognimaux and Lemoine 1983

MOD06

HH0=m+n*SS0+o*logSS0

Logarithmic equation

Newland 1989

MOD07

HH0=m+expn*SS0

Exponential equation

Elagib and Mansell 2000

MOD08

HH0=m+n(SS0)c

Hybrid equation

Elagib and Mansell 2000

MOD09

HH0=m++oZ+pSS0

Hybrid equation

Elagib and Mansell 2000

MOD10

HH0=m+nZ+oSS0

Hybrid equation

Elagib and Mansell 2000

MOD11

HH0=m+ncosφ+oSS0

Linear equation

Raja and Twidell 1990

MOD12

HH0=m(T)0.5

Power equation

Allen 1997

MOD13

HH0=m+n(T)0.5

Hybrid equation

Hargreaves 1985

MOD14

HH0=m1-exp-nTo

Hybrid equation

Bristow and Campbell 1984

MOD15

HH0=m+n*lnT

Logarithmic equation

Chen et al. 2004

MOD16

H=m+nSS0+osinδ+pTmax

Linear equation

Chen et al. 2004

MOD17

H=m+nH0+oSS0+psinδ+qTmax+rRH

Linear equation

Chen et al. 2004

MOD18

H=m+nH0+oSS0+pRH+qST+rTmax

Linear equation

Chen et al. 2004

MOD19

H=m+nH0+oSS0+psinδ+qRH+rST+sTmax

Linear equation

Chen et al. 2004

MOD20

H=m+nH0++pRH+qSS0+rT+sST+tP

Linear equation

Ertekin and Yaldiz 1999

MOD21

HH0=m+nSS0+oTmax+pRH+qTmax(SS0)

Linear equation

Ododo et al.1995

MOD22

H=m+nH0+oTmax+pTmin+qV

Linear equation

El-Metwally 2004

MOD23

H=m+nSS0+osinδ+pT

Linear equation

Togrul and Onat 1999

MOD24

H=m+nH0+oSS0+psinδ+qT+rRH

Linear equation

Togrul and Onat 1999

MOD25

H=m+nSS0+osinδ+pT+qRH

Linear equation

Togrul and Onat 1999

MOD26

H=m+nH0+oSS0+pST+qRH

Linear equation

Togrul and Onat 1999

MOD27

H=m+nH0+oSS0+pRH+qST+rT

Linear equation

Togrul and Onat 1999

MOD28

H=m+nH0+oSS0+psinδ+qT+rST+sRH

Linear equation

Togrul and Onat 1999

MOD29

H=mSS0nRH0

Power equation

Swartzman and Ogunlade 1967

MOD30

HH0=m+nSS0+oRH

Linear equation

Swartzman and Ogunlade 1967

MOD31

HH0=m+nSS0+oW

Hybrid equation

Garg and garg 1982

MOD32

HH0=m++oW

Hybrid equation

Garg and garg 1982

MOD33

HH0=m(T)n (1+o*P+pP2)

Power equation

De Jong and Stewart 1993

MOD34

H=m+n(T)0.5H0+oTmax+pP+qP2

Hybrid equation

Hunt et al. 1998

MOD35

H=m+nH0+oSS0+pRH+qTmax+rsinδ

Linear equation

Coulibaly and Ouedraogo 2016

Here RH, Tmax and Tmin are respectively the mean monthly relative humidity (in percentage), maximum and minimum air temperature (°C), W(cm) is the atmospheric precipitable water vapor per unit volume of air (cm) computed according to Leckner 1978 , H is the measured mean monthly global solar radiation, H0 the computed monthly average of daily extraterrestrial radiation, S the effective day length, S0 the computed maximum possible sunshine duration, P the precipitation in (mm), 𝑆𝑇 is the mean soil temperature (°C), T=(Tmax-Tmin) is the temperature difference (°C), Z is the Altitude (Km) and C the cloudiness (cloud cover)
W=0.0049RHexp(26.23-5416/Tk)Tk(1)
Tk is the air temperature (in Kelvin).
H0=24*Gscπ1+0,033cos360n365(cosϕcosδsinωs+πωs180sinϕsinδ)in Wh/m2(2)
δ=23.45sin360365(284+n)(3)
ωs=cos-1-tanϕtanδ(4)
S0=215ωs(5)
Gsc= is the solar constant (1367 W/m2), ϕ=latitude (deg), n= day of year 1n365, δ is the declination (deg) and ωs is the hour angle (deg)
2.2. Evaluation Parameters of the Model Performance
In the present study statistical performance parameters, mentioned are used the strength of models. The RMSE measures the average difference between a statistical model’s predicted values and the measured values. Mathematically, it represents the distance between the regression line and the data points. the RMSE is always positive. Low RMSE values indicate that the model fits the data well and has more precise predictions. A zero value is ideal. Mean percentage error (MPE), is described as the measure of the extent of the error of values in terms of percentage of the observed or measured value, The MBE therefore evaluate underestimation and over estimation, underestimation results in a negative value of MBE while a positive value represents an overestimation. MBE has a low desirable value and ideally its value should be zero. One drawback of this test is that over-estimation of an individual observation will cancel under-estimation in a separate observation . The coefficient of determination R2 determine how well the regression line approximates the real data points. R2 range between 0 and 1 ideal value is 1 which means the best goodness of fit of model. Statistical parameters are defined as follows in table 5.
Table 5. Performance metrics for model evaluations.

parameters

equations

Root mean squared error (RMSE)

1ni=1n(Yi,m-Yi,c)21/2

Normalised Root mean squared error RMSE (%)

RMSE Y̅m*100

Mean bias error (MBE)

1ni=1n(Yi,c-Yi,m) 

Normalised Mean absolute bias error MBE (%)

MBE Y̅m*100

Mean percentage error MPE (%)

1ni=1n(Yi,m-Yi,cYi,m)*100

Determination coefficient (R2)

1-i=1n(Yi,m-Yi,c)2i=1n(Yi,m-Y̅m)2

These metrics are preferred to comparing the predictive performance of the models over different datasets, where Yi,m is the ith measured data, Yi,c is the ith calculated data, Y̅m is is the mean of the measured values and n is the total number of the observations.
3. Results and Discussion
3.1. Performance Statistics of Models
For the two cities of Garoua and Maroua models are compared using regression analysis by considering mentioned statistical parameters MBE, MPE, RMSE, R2, and their associated ranking as shown in table 6 and table 7 respectively for Garoua and Maroua. these tables present informations on the accuracy of each model involved. Through these informations one can be able to select the best model for a specific application. indeed, from these data tables it is easily seen that the MBE (%), which is the metering of underestimation (negative data) and overestimation (positive data) with respect to the measured ones, lies between -0.269% and +0.162% for the city of Maroua; -1.41% and +0.571% for the city of Garoua. Likewise, the RMSE (%), lies between 0.862% and 6.343% for the city of Maroua; 0.809% and 13.637% for the city of Garoua. (R2) however, lies between 0.15 to 0.985 for Maroua and between 0.04 to 0.958 for Garoua. Regarding MPE (%) the predicted values are between - 0.00845 and -0.43614 for the site of Maroua, -0.00631 and 1.45545 for Garoua. Indeed, when we consider MBE (%) as accuracy criteria, the most performant models are: Tugrul and Onat 1999(MOD27) for Maroua (MBE =1,34E-15% ), Tugrul and Onat 1999 (MOD23) for Garoua,(MBE= 1,32E-15%). However, taking into account RMSE, MPE and R2, the most performant models are: Ertekin and Yaldiz 1999 (MOD20) for Maroua (RMSE=0,86221%, MPE=-0.00845, R2=0,98522), Tugrul and Onat 1999 (MOD28) for Garoua (RMSE=0,80631%, MPE=-0.00631, R2=0,95934).
Table 6. Statistical parameters comparison for the city of Maroua with their associated ranking (+is overestimation and –is under estimation).

Models

MBE (%)

Ranking

RMSE (%)

Ranking

R2

Ranking

MPE (%)

Ranking

Statute

Number of Variables

MOD01

1.62E-02

18

5.39206

29

0.42207

29

-0.30831

29

+

3

MOD02

1.62E-02

20

5.39206

31

0.42207

31

-0.30831

31

+

4

MOD03

3.30E-02

25

3.79059

17

0.71438

17

-0.14033

16

+

5

MOD04

2.28E-02

23

5.28211

25

0.44540

25

-0.29549

26

+

3

MOD05

2.23E-02

22

5.39208

32

0.42206

32

-0.31439

32

+

6

MOD06

5.47E-02

32

4.06724

19

0.67117

19

-0.16673

18

+

4

MOD07

8.30E-03

15

5.39716

34

0.42098

34

-0.30176

27

+

3

MOD08

1.62E-01

34

5.28700

26

0.44437

26

-0.43614

35

+

3

MOD09

1.62E-02

19

5.39206

30

0.42207

30

-0.30831

30

+

5

MOD10

-3.88E-02

26

5.39204

28

0.42207

28

-0.25306

24

-

4

MOD11

2.59E-02

24

5.39212

33

0.42206

33

-0.31805

33

+

4

MOD12

-2.69E-01

35

6.54365

35

0.14885

35

0.39683

34

-

3

MOD13

5.35E-02

30

4.47961

22

0.60111

22

-0.21607

22

+

3

MOD14

6.03E-02

33

3.97852

18

0.68536

18

-0.17544

19

+

3

MOD15

5.44E-02

31

4.42747

21

0.61035

21

-0.21081

21

+

3

MOD16

-1.75E-14

5

1.94234

12

0.92501

12

-0.03617

10

-

4

MOD17

-3.09E-14

11

1.45003

5

0.95821

5

-0.02259

4

-

6

MOD18

-2.42E-14

7

1.69333

9

0.94300

9

-0.03134

9

-

6

MOD19

-3.23E-14

13

1.45003

4

0.95821

4

-0.02259

6

-

7

MOD20

-2.82E-14

8

0.86221

1

0.98522

1

-0.00845

1

-

8

MOD21

4.80E-02

29

2.46654

15

0.87907

15

-0.06946

15

+

6

MOD22

-3.90E-14

14

1.92510

10

0.92633

10

-0.04064

13

-

5

MOD23

-1.48E-14

4

1.97613

14

0.92238

14

-0.03671

12

-

4

MOD24

-1.88E-14

6

1.30978

3

0.96590

3

-0.01845

3

-

6

MOD25

-3.09E-14

12

1.93499

11

0.92557

11

-0.03639

11

-

5

MOD26

-1.34E-14

3

1.96217

13

0.92347

13

-0.04378

14

-

5

MOD27

1.34E-15

1

1.61230

7

0.94833

7

-0.02827

8

+

6

MOD28

1.34E-15

2

1.23958

2

0.96946

2

-0.01581

2

+

7

MOD29

1.25E-02

17

5.08758

24

0.48550

24

-0.28579

25

+

3

MOD30

3.91E-02

27

4.34836

20

0.62415

20

-0.20995

20

+

4

MOD31

1.86E-02

21

5.36998

27

0.42679

27

-0.30772

28

+

4

MOD32

-1.14E-02

16

4.60543

23

0.57839

23

-0.23276

23

-

4

MOD33

4.69E-02

28

3.54179

16

0.75065

16

-0.14315

17

+

4

MOD34

-2.82E-14

9

1.61718

8

0.94801

8

-0.02477

7

-

5

MOD35

-2.82E-14

10

1.45003

6

0.95821

6

-0.02259

5

-

6

Table 7. Statistical parameters comparison for the city of Garoua with their associated ranking (+is overestimation and –is under estimation).

Models

MBE (%)

Ranking

RMSE (%)

Ranking

R2

Ranking

MPE (%)

Ranking

Statute

Number of variables

MOD01

6.80E-02

25

4.10866

27

-0.05574

25

-0.17394

28

+

3

MOD02

6.80E-02

24

4.10866

26

-0.05574

27

-0.17394

27

+

4

MOD03

7.70E-02

31

3.63587

18

0.17326

18

-0.13255

19

+

5

MOD04

7.08E-02

28

4.12084

32

-0.06200

22

-0.17526

31

+

3

MOD05

-5.32E-01

33

4.13393

33

-0.06876

21

0.42599

33

-

6

MOD06

8.02E-02

32

3.63834

19

0.17213

19

-0.13331

20

+

4

MOD07

7.46E-02

29

4.10903

29

0.01440

33

-0.18054

32

+

3

MOD08

5.71E-01

34

4.16641

34

0.00520

34

-0.67415

34

+

3

MOD09

6.80E-02

26

4.10866

28

-0.05574

26

-0.17394

29

+

5

MOD10

6.26E-02

21

4.10849

25

-0.05565

28

-0.16855

24

+

4

MOD11

6.10E-02

19

4.10844

24

-0.05563

29

-0.16696

23

+

4

MOD12

-1.41E+00

35

13.63798

35

NaN

/

1.45545

35

-

3

MOD13

6.72E-02

23

4.10462

23

-0.05366

30

-0.17324

25

+

3

MOD14

4.12E-02

18

3.90222

22

0.04769

32

-0.11025

16

+

3

MOD15

6.87E-02

27

4.11402

31

-0.05850

23

-0.17425

30

+

3

MOD16

-1.84E-14

8

1.53924

9

0.85183

9

-0.02377

9

-

4

MOD17

-2.50E-14

12

0.99261

4

0.93838

4

-0.00954

5

-

6

MOD18

-1.32E-14

6

2.39040

13

0.64265

12

-0.05881

13

-

6

MOD19

-6.58E-15

2

0.97779

3

0.94021

3

-0.00933

3

-

7

MOD20

-2.10E-14

10

1.51477

8

0.85650

8

-0.02351

8

-

8

MOD21

7.63E-02

30

3.45200

16

0.25476

16

-0.12084

17

+

6

MOD22

-6.58E-15

3

2.27739

12

0.67564

12

-0.05371

12

-

5

MOD23

1.32E-15

1

1.23965

7

0.90389

7

-0.01547

7

+

4

MOD24

1.18E-14

5

0.80925

2

0.95904

2

-0.00637

2

+

6

MOD25

-1.97E-14

9

1.07212

6

0.92811

6

-0.01127

6

-

5

MOD26

-2.24E-14

11

2.41242

14

0.63603

14

-0.06002

14

-

5

MOD27

2.76E-14

13

2.08012

11

0.72940

11

-0.04461

11

+

6

MOD28

-9.21E-15

4

0.80631

1

0.95934

1

-0.00631

1

-

7

MOD29

-3.99E-03

15

3.69038

20

0.14828

20

-0.13383

21

-

3

MOD30

6.69E-02

22

4.10971

30

-0.05628

24

-0.17380

26

+

4

MOD31

6.15E-02

20

3.89063

21

0.05334

31

-0.15382

22

+

4

MOD32

2.65E-02

16

3.46197

17

0.25045

17

-0.12182

12

+

4

MOD33

3.79E-02

17

2.58807

15

0.58110

15

-0.06547

15

+

4

MOD34

-1.32E-14

7

1.98244

10

0.75421

10

-0.03996

10

-

5

MOD35

-3.29E-14

14

0.99261

5

0.93838

5

-0.00954

4

-

6

According to performance criteria, it appears that most of the models provide good performance since values of statistical parameters obey to performance criterion. Indeed, -5% < MBE < +5% and RMSE(%) is less than 15%. This shows in general that models could be helpful for the prediction of global solar irradiation in each city. In fact, goodness of the model associate to their ranking are important since they show how accurate the data are. From these results we can notice that models which more detailed atmospheric information fulfill performance better than those with less or no such inputs. Thus two criteria can be retained for models performance evaluation (1) the best models according MBE criterion (RMSE and MPE are fulfilled) (2) the best model according to RMSE, MPE and R2 criteria. However, for the most accurate model’s selection, criteria according to RMSE, MPE and R2 is more significant. The reports are shown in the Table 8.
Table 8. Best models according to two criteria for each city.

City

Ranking

MBE criterion Best model

Authors

RMSE and R2 criterion Best model

Authors

Maroua

1

MOD27

Togrul and Onat 1999

MOD20

Ertekin and Yaldiz 1999

2

MOD28

Togrul and Onat 1999

MOD28

Togrul and Onat 1999

Garoua

1

MOD23

Togrul and Onat 1999

MOD28

Togrul and Onat 1999

2

MOD19

Chen et al. 2004

MOD24

Togrul and Onat 1999

3.2. Regression’s Coefficients of Models
In order to help experienced solar radiation developer, engineers as well as new comer, regression coefficient for different models are presented in Table 9 and Table 10 respectively for different cities.
Table 9. Coefficients of Regression for models (city of Maroua).

Regression coefficients

Models

m

n

o

p

q

r

s

t

MOD01

0,50195

0,08354

-

-

-

-

-

-

MOD02

-0,99476

0,08354

-

-

-

-

-

-

MOD03

-2,45841

12,86417

-17,94068

8,21964

-

-

-

-

MOD04

0,58445

0,14654

-

-

-

-

-

-

MOD05

0,63980

-0,07334

-0,01317

0,85120

-

-

-

-

MOD06

2,40452

-1,92038

3,07752

-

-

-

-

-

MOD07

-0,49590

0,07812

-

-

-

-

-

-

MOD08

-4,24200

4,82700

0,01311

-

-

-

-

-

MOD09

0,50195

0,00000

0,00000

0,08354

-

-

-

-

MOD10

-3,61500

9,73200

0,08354

-

-

-

-

-

MOD11

0,81550

0,62130

0,08354

-

-

-

-

-

MOD12

0,16281

-

-

-

-

-

-

-

MOD13

0,32766

0,06764

-

-

-

-

-

-

MOD14

0,57190

0,00313

3,07600

-

-

-

-

-

MOD15

0,28045

0,26159

-

-

-

-

-

-

MOD16

0,92146

1,30876

-0,07061

0,10798

-

-

-

-

MOD17

-0,46146

0,23174

0,77475

-0,09324

0,09453

-0,00161

-

-

MOD18

-0,03309

0,26264

0,64479

-0,00746

-0,06859

0,14336

-

-

MOD19

-0,46494

0,23126

0,77677

-0,09332

-0,00156

0,00056

0,09417

-

MOD20

-2,83785

0,51644

0,02808

0,00732

2,42812

-0,05528

0,10088

-0,00256

MOD21

-0,28335

0,82476

0,02429

-0,00007

-0,02346

-

-

-

MOD22

-0,36681

0,20886

0,16381

-0,07969

-

-

-

-

MOD23

0,67449

2,02037

-0,08645

0,12187

-

-

-

-

MOD24

-0,85031

0,25937

1,31109

-0,10799

0,10580

-0,00193

-

-

MOD25

1,27373

1,62879

-0,08461

0,11458

-0,00264

-

-

-

MOD26

-1,21914

0,06755

1,39162

0,15549

0,01040

-

-

-

MOD27

-0,23736

0,38164

1,42327

-0,01496

-0,14584

0,21356

-

-

MOD28

-0,39010

0,38106

1,28143

-0,10351

0,18635

-0,11602

-0,01198

-

MOD29

8,33400

-0,02159

-0,11270

-

-

-

-

-

MOD30

0,73035

-0,13818

-0,00162

-

-

-

-

-

MOD31

0,52984

0,05641

-0,00312

-

-

-

-

-

MOD32

0,64045

0,00175

-0,02710

-

-

-

-

-

MOD33

1,21900

-0,29230

-0,00060

0,00000

-

-

-

-

MOD34

0,29359

0,07710

0,07405

0,00311

-0,00001

-

-

-

MOD35

-0,46146

0,23174

0,77475

-0,00161

0,09453

-0,09324

-

-

Table 10. Coefficients of regression for the models (city of Garoua).

Regression coefficients

Models

m

n

o

p

q

r

s

t

MOD01

0,58530

-0,02287

-

-

-

-

-

-

MOD02

-0,58776

-0,02287

-

-

-

-

-

-

MOD03

-0,38856

4,00975

-5,34915

2,28116

-

-

-

-

MOD04

0,56572

-0,02265

-

-

-

-

-

-

MOD05

-14,50000

2,61600

1,61600

-24,44000

-

-

-

-

MOD06

1,58615

-1,08172

1,59972

-

-

-

-

-

MOD07

-0,41470

-0,02300

-

-

-

-

-

-

MOD08

-13,08000

13,65000

-0,00045

-

-

-

-

-

MOD09

0,58530

0,00000

0,00000

-0,02287

-

-

-

-

MOD10

0,39370

0,79490

-0,02287

-

-

-

-

-

MOD11

0,59190

0,00667

-0,02287

-

-

-

-

-

MOD12

0,16277

-

-

-

-

-

-

-

MOD13

0,59559

-0,00755

-

-

-

-

-

-

MOD14

0,57230

0,00000

8,15800

-

-

-

-

-

MOD15

0,59543

-0,02423

-

-

-

-

-

-

MOD16

3,97413

-0,41133

-0,19924

0,05552

-

-

-

-

MOD17

1,92743

0,14616

-0,10226

-0,21833

0,05933

0,00451

-

-

MOD18

1,98090

0,05794

0,05706

0,00547

0,04372

0,04263

-

-

MOD19

1,72338

0,12160

-0,10084

-0,21818

0,00629

0,03927

0,03737

-

MOD20

-0,74307

0,78370

0,04067

0,00256

1,63266

-0,08328

0,00530

-0,00474

MOD21

-0,21992

0,95016

0,02375

0,00030

-0,02934

-

-

-

MOD22

2,65880

0,08881

0,04149

0,02815

-

-

-

-

MOD23

3,75063

0,11129

-0,18481

0,06256

-

-

-

-

MOD24

2,12299

0,13217

0,32632

-0,20191

0,06221

0,00323

-

-

MOD25

3,21838

0,44816

-0,18992

0,06750

0,00280

-

-

-

MOD26

1,63268

0,01509

0,06074

0,11705

0,00867

-

-

-

MOD27

3,11028

0,12989

0,93031

-0,00377

-0,10492

0,13117

-

-

MOD28

2,22119

0,13934

0,37764

-0,20013

0,06971

-0,01330

0,00245

-

MOD29

6,09100

0,05135

-0,01490

-

-

-

-

-

MOD30

0,57573

-0,01439

0,00007

-

-

-

-

-

MOD31

0,50463

0,04478

0,00938

-

-

-

-

-

MOD32

0,58757

0,00110

-0,00495

-

-

-

-

-

MOD33

0,46820

0,06392

0,00190

-0,00001

-

-

-

-

MOD34

2,06496

0,04671

0,04757

0,00895

-0,00003

-

-

-

MOD35

1,92743

0,14616

-0,10226

0,00451

0,05933

-0,21833

-

-

3.3. Comparison of the Best Predicted Models with Measured and Satellite Derived Data
Nowadays different databases are used by solar energy planners and engineers for the designing of PV solar power in the absence of measured data. These databases are sometimes used in developing countries due to the lack of measured data. To better understand the behavior of predicted data, Figure 1 and Figure 2, respectively for the city of Maroua and Garoua, compared these data.
From above figures, we can see best models predict the trend of the measured global solar radiation in these regions indeed, there is no visible differences between measured and predicted data from best models named BestModel1MOD28, Bestmodel2MOD23 for the city of Garoua, and BestModel1MOD20, BestModel2MOD27 for Garoua. nevertheless, using comparison of predicted data with others resources data like Retscreen, Solargis, and PVgis which are commonly used for energy planning and management, we can easily make decision on how projects are overdesigned or under designed according to real dataset. Overestimation and underestimation are presented in Table 11.
Table 11. Comparison of others solar resources with best predicted model data.

Maroua city

Garoua city

Month

Retscreen

pvgis

Solargis

Retscreen

pvgis

Solargis

January

over-estimation

over-estimation

over-estimation

over-estimation

over-estimation

over-estimation

February

over-estimation

over-estimation

over-estimation

over-estimation

over-estimation

over-estimation

March

over-estimation

over-estimation

over-estimation

over-estimation

over-estimation

over-estimation

April

over-estimation

over-estimation

over-estimation

over-estimation

over-estimation

over-estimation

May

over-estimation

over-estimation

over-estimation

under-estimation

over-estimation

over-estimation

June

over-estimation

over-estimation

over-estimation

under-estimation

over-estimation

under-estimation

July

over-estimation

over-estimation

over-estimation

under-estimation

under-estimation

under-estimation

August

over-estimation

over-estimation

over-estimation

under-estimation

under-estimation

under-estimation

September

over-estimation

over-estimation

over-estimation

under-estimation

under-estimation

under-estimation

October

over-estimation

over-estimation

over-estimation

under-estimation

over-estimation

under-estimation

November

over-estimation

over-estimation

over-estimation

over-estimation

over-estimation

over-estimation

December

over-estimation

under-estimation

over-estimation

over-estimation

over-estimation

over-estimation

Figure 1. Others resources data and best models for the city of Maroua.
Figure 2. Others resources data and best models and for the city of Garoua.
4. Conclusion
Solar radiation resources data are one of the keys to energy projects success. This research was conducted to evaluate performance of some representative empirical models encountered in the open literature. The comparison is made possible using statistical evaluation parameters deduced from regression analysis realized using Excel and Matlab tools. Amongst the thirty-five (35) empirical models studied, accuracy of the models was verified by comparing estimated values with measured values in terms of the following statistical evaluation parameters named root mean square error (RMSE, mean bias error (MBE),), and the determination coefficient (R2). It is observed that more meteorological data are needed for the precise evaluation of the global solar radiation. The results shows that the models of Togrul and Onat 1999 (MOD28), Ertekin and Yaldiz 1999 (MOD20) appear to be more accurate and performed data better. Through the results obtained we clearly demonstrate that formulated models are good enough to be used to predict monthly average daily radiation for these two cities in Cameroon. It may be concluded that the models presented in this study could be used to estimate accurately the solar radiation at any semi-arid region around the world.
Abbreviations

BSh

Semi-Arid Climate Zone

Gsc

Solar Constant (W/m2)

H0

Extraterrestrial Solar Radiation (kWh/m2)

H

Measured Solar Radiation (kWh/m2)

MBE

Mean Bias Error (kWh/m2)

MPE

Mean Percentage Error (kWh/m2)

P 

Precipitation in (mm)

PV

Photovoltaic

RH

Relative Humidity in Percentage

RMSE

Root Mean Square Error (kWh/m2)

R2

Determination Coefficient

S

Effective Sunshine Duration (h)

S0

Day Length (h)

ST

Mean Soil Temperature (°C)

T

Monthly Mean Temperature (°C)

Tk

Monthly Daily Mean Air Temperature (K.)

Tmax

Mean Maximum Temperature (°C)

Tmin

Mean Minimum Temperature (°C)

W

Precipitable Water Vapor from the Atmosphere (cm).

Y̅m

Mean Annual Solar Radiation (kWh/m2)

Yi,c

Calculated Solar Radiation (kWh/m2)

Yi,m

Measured Solar Radiation (kWh/m2)

Z

Altitude (Km)

Acknowledgments
The authors of this manuscript are thankful to the National Advanced school of Engineering of the university of Maroua and Agency for the Safety of Air Navigation in Africa (ASECNA) for providing data which permit to carry out this article.
Author Contributions
Kodji Deli: Conceptualization, Data curation, Formal Analysis, Investigation, Methodology, Resources, Software, Supervision, Validation, Visualization, Writing – original draft, Writing – review & editing
Etienne Tchoffo Houdji: Data curation, Formal Analysis, Investigation, Methodology, Software, Validation, Visualization, Writing – original draft
Albert Ayang: Data curation, Formal Analysis, Investigation, Methodology, Software, Validation, Visualization, Writing – original draft
Dieudonne Kidmo Kaoga: Data curation, Formal Analysis, Investigation, Methodology, Software, Supervision, Validation, Visualization, Writing – original draft
Noel Djongyang: Conceptualization, Data curation, Formal Analysis, Investigation, Methodology, Software, Supervision, Validation, Visualization, Writing – original draft
Funding
Authors declare that no funding for research has been received from any funding agency.
Conflicts of Interest
The authors declare no conflicts of interest.
References
[1] Ayodele, T. R., Ogunjuyigbe, A. S. O. Performance assessment of empirical models for prediction of daily and monthly average global solar radiation: the case study of Ibadan, Nigeria. International Journal of Ambient Energy. 2016;
[2] Badescu, V. Modeling Solar Radiation at the Earth’s Surface, Springer-Verlag Berlin Heidelberg. 2008 ISBN: 978-3-540-77454-9.
[3] Katiyar, A. K., Pandey, C. K. Simple correlation for estimating the global solar radiation on horizontal surfaces in India. Energy. 35 (2010), pp. 5043–5048.
[4] Liu, D. L., Scott, B. J. Estimation of solar radiation in Australia from rainfall and temperature observations, Agriculture and Forest Meteorology.; 106(2001) 41–59.
[5] Trnka M., Zalud, Z., Eitzinger, J., Dubrovsky, M. Global solar radiation in Central European lowlands estimated by various empirical formulae, Agriculture and Forest Meteorology. 131(2005), pp. 54–76.
[6] Menges, H. O., Ertekin, C., Sonmete, M. H. Evaluation of global solar radiation models for Konya Turkey, Energy Conversion and Management. 47 (2006), pp. 3149–3173.
[7] Ertekin, C., Evrendilek, F. Spatio-temporal modeling of global solar radiation dynamics as a function of sunshine duration for Turkey, Agriculture and Forest Meteorology. 145 (2007), 36–47.
[8] Evrendilek, F., Ertekin, C. Assessing solar radiation models using multiple variables over Turkey, Climate Dynamics, 31(2008) 131–149.
[9] Sonmete, M. H., Ertekin, C., Menges, H. O., Hacıseferogullari, H., Evrendilek, F. Assessing monthly average solar radiation models: a comparative case study in Turkey, Environ Monit Assess; 175(2011) 251–277.
[10] Angstrom, A. Solar and terrestrial radiation, Quarterly Journal of Royal Meteorological Society, 50(1924) 121–125.
[11] Prescott, J. A. Evaporation from water surface in relation to solar radiation, Transactions of the Royal Society of Australia. 46 (1940) 114–118.
[12] Glower, J., McGulloch, J. S. G. The empirical relation between solar radiation and hours of sunshine, Quart J R Met Soc. 84(1958) 172-175.
[13] Dogniaux, R., Lemoine, M. Classification of radiation sites in terms of different indices of atmospheric transparency, Solar energy research and development in the European Community Series F. 2(1983). Dordrecht: Reidel.
[14] Raja, I. A., Twidell, J. W. Distribution of global insolation over Pakistan. Solar Energy. 44(1990) 63–71.
[15] Samuel, T. D. M. A. Estimation of global radiation for Srilanka, Solar Energy. 47(5) (1991) 333-337.
[16] Ampratwum, D. B., Dorvio, A. S. S. Estimation of solar radiation from the number of sunshine hours, Applied Energy. 63(1999) 161–167.
[17] Ayangma, F., Nkeng, G. E., Bonoma, D. B., Nganhou, J. Evaluation du potentiel solaire au Cameroun: cas du Nord Cameroun, African Journal of Science and Technology. 9 (2) (2008) 32-40.
[18] Okundamiya, M. S., Emagbetere, J. O., Ogujor, E. A. Evaluation of various global solar radiation models for Nigeria, International Journal of Green Energy; 13(5) (2016) 505-512.
[19] Ertekin, C., Yaldiz, O. Comparison of some existing models for estimating global solar radiation for Antalya (Turkey), Energy Conversion & Management. 41(2000) 311-330.
[20] Mboumboue, E., Njomo, D., Ndiaye, M. L., N'diaye, P. A., Ndiaye, M. F. Tossa, A. K On the applicability of several conventional regression models for the estimation of solar global radiation component in Cameroon and Senegal sub-Saharan tropical regions, Journal of Renewable and Sustainable Energy. 8(2016), 025906;
[21] Afungchui, D. J., Ebobenow, N. R., Ngwa, N. A. Global Solar Radiation of some Regions of Cameroon using the Linear Angstrom Model and Non-linear Polynomial Relations: Part 2, Sun-path Diagrams, Energy Potential Predictions and Statistical Validation. International Journal of Renewable Energy Research. 8(1) (2018).
[22] Wild, M. Global dimming and brightening: A review, Journal of Geophysical Research., 114(2009) D00D16.
[23] Cebecauer, T., Suri, M. Typical Meteorological Year data: SolarGIS approach, Energy Procedia. 69 (2015) 1958 – 1969.
[24] Ineichen, P., Barroso, C. S., Geiger, B., Hollmann, R. A. Marsouin, R. Müller, Satellite Application Facilities irradiance products: hourly time step comparison and validation over Europe, International Journal of Remote Sensing.; 30 (2009) 5549-5571.
[25] Müller, R., Matsoukas, C., Gratzki, A., Behr, H. D., Hollmann, R. The CM-SAF operational scheme for the satellite based retrieval of solar surface irradiance - A LUT based eigenvector hybrid approach. Remote Sensing of Environment.; 113 (2009) 1012-1024.
[26] Leckner, B. The spectral distribution of solar radiation at the earth’s surface-elements of a model. Solar Energy. 20 (1978) 143–150.
[27] Page, J. K. The estimation of monthly mean values of daily total short wave radiation on vertical and inclined surfaces from sunshine records for latitudes 40°N–40°S. In Proceedings of UN conference on new sources of energy. (1961) 378–390.
[28] Newland, F. J. A study of solar radiation models for the coastal region of South China, Solar Energy. 31(1988) 227–235.
[29] Elagib, N., Mansell, M. G. New approaches for estimating global solar radiation across Sudan, Energy Conversion and Management.; 41(2000) 419–434.
[30] Allen, R. Self calibrating method for estimating solar radiation from air temperature. Journal of Hydrologic Engineering. 2(1997) 56–67.
[31] Hargreaves, G. L., Hargreaves, G. H. Riley, P. Irrigation water requirement for the Senegal River Basin. Journal of Irrigation and Drainage Engineering ASCE. 111(1985) 265–275.
[32] Bristow, K. L., Campbell, G. S. On the relationship between incoming solar Radiation and daily maximum and minimum temperature. Agricultural and Forest Meteorology. 31 (1984) 159-166.
[33] Chen, R., Ersi, K., Yang, J., Lu, S., Zhao, W. Validation of five global radiation models with measured daily data in China, Energy Conversion and Management. 45(2004) 1759–1769.
[34] Ertekin, C., Yaldiz, O. Estimation of monthly average daily global radiation on horizontal surface for Antalya, Turkey, Renewable Energy. 17(1) (1999) 95-102.
[35] Ododo, J. C., Sulaiman, A. T., Aidan, J., Yuguda, M. M., Ogbu, F. A. The importance of maximum air temperature in the parameterisation of solar radiation in Nigeria, Renewable Energy. 6 (1995) 751–763.
[36] El-Metwally. M. Simple new methods to estimate global solar radiation based on meteorological data in Egypt, Atmospheric Research. 69(2004) 217–239. (Article)
[37] Togrul, I. T., Onat, E. A study for estimating solar radiation in Elazig using geographical and meteorological data, Energy Conversion and Management; 40(1999) 1577–1584.
[38] Swartman, R. K., Ogunlade, O. Solar radiation estimates from common parameters. Solar Energy; 11(1967) 170–172.
[39] Garg, H. P., Garg, S. T. Prediction of global solar radiation from bright sunshine hours and other meteorological Data, Energy conversion management.; 23(2) (1982) 113-118.
[40] DeJong, R., Stewart, D. W. Estimating global solar radiation from common meteorological observations in western Canada, Canadian Journal of Plant Science. 73(1993) 509-518.
[41] Hunt, L. A., Kuchar, L., Swanton, C. J. Estimation of solar radiation for use in crop modelling, Agricultural And Forest Meteorology. 91(1998) 293-300.
[42] Coulibaly, O., Ouedraogo. A. Correlation of Global Solar Radiation of Eight Synoptic Stations in Burkina Faso Based on Linear and Multiple Linear Regression Methods Hindawi Publishing Corporation Journal of Solar Energy. (2016). ID 7870907, 9 pages
[43] Ahmad, M. J., Tiwari, G. N. Evaluation and comparison of hourly solar radiation models, Int. J. Energy Res.; 33(2009) 538–552.
[44] Akinoglu, B. G. Recent Advances in the Relations between Bright Sunshine Hours and Solar. In: Badescu V, editor. Modeling solar radiation at the earth surface. Berlin: Springer. (2008) chapter 5: 115–143.
Cite This Article
  • APA Style

    Deli, K., Houdji, E. T., Ayang, A., Kaoga, D. K., Djongyang, N. (2024). Performance Comparison of Empirical Models for Estimating Global Solar Irradiation in the Soudano-Sahelian Zone of Cameroon: The Case of the City of Maroua and Garoua. International Journal of Sustainable and Green Energy, 13(2), 28-42. https://doi.org/10.11648/j.ijrse.20241302.12

    Copy | Download

    ACS Style

    Deli, K.; Houdji, E. T.; Ayang, A.; Kaoga, D. K.; Djongyang, N. Performance Comparison of Empirical Models for Estimating Global Solar Irradiation in the Soudano-Sahelian Zone of Cameroon: The Case of the City of Maroua and Garoua. Int. J. Sustain. Green Energy 2024, 13(2), 28-42. doi: 10.11648/j.ijrse.20241302.12

    Copy | Download

    AMA Style

    Deli K, Houdji ET, Ayang A, Kaoga DK, Djongyang N. Performance Comparison of Empirical Models for Estimating Global Solar Irradiation in the Soudano-Sahelian Zone of Cameroon: The Case of the City of Maroua and Garoua. Int J Sustain Green Energy. 2024;13(2):28-42. doi: 10.11648/j.ijrse.20241302.12

    Copy | Download

  • @article{10.11648/j.ijrse.20241302.12,
      author = {Kodji Deli and Etienne Tchoffo Houdji and Albert Ayang and Dieudonne Kidmo Kaoga and Noel Djongyang},
      title = {Performance Comparison of Empirical Models for Estimating Global Solar Irradiation in the Soudano-Sahelian Zone of Cameroon: The Case of the City of Maroua and Garoua
    },
      journal = {International Journal of Sustainable and Green Energy},
      volume = {13},
      number = {2},
      pages = {28-42},
      doi = {10.11648/j.ijrse.20241302.12},
      url = {https://doi.org/10.11648/j.ijrse.20241302.12},
      eprint = {https://article.sciencepublishinggroup.com/pdf/10.11648.j.ijrse.20241302.12},
      abstract = {The main objective of this study is to compare thirty-five (35) solar radiation models available in the open literature in order to predict monthly solar radiation in two main cities of Cameroon. This estimation and comparison are based on selected statistical comparison parameters named, root mean square error (RMSE), mean bias error (MBE), mean percentage error (MPE) and determination coefficient (R2). These different models are implemented using regression analysis tools named Exel and MATLAB. Estimated values were compared with measured values according to normalized values of statistical parameters, using measured meteorological data of more than 19 years, from 1984 to 2015. All the models have been classified with their associated ranking according to their statistical parameter accuracy. From this study it appears that the models of Ertekin and Yaldiz  (MOD20), Togrul and Onat (MOD28), are more accurate than other models. Indeed, for the city of Maroua (MBE%=-2.82E-14; RMSE%=0.862; MPE=-0.00845; R2=0.985), while for Garoua (MBE%=-9.21E-15; RMSE%=0.806; MPE=-0.00631; R2=0.959). according to their accuracy these models can be therefore be used to predict monthly solar radiation for soudano-sahelian regions of Cameroon. Correlation equations found in this paper will help solar energy researcher to estimate data with trust because of its fine agreement with the observed one. hence the models presented in this study could be used to evaluate accurately the solar radiation at any locations with similar climate.
    },
     year = {2024}
    }
    

    Copy | Download

  • TY  - JOUR
    T1  - Performance Comparison of Empirical Models for Estimating Global Solar Irradiation in the Soudano-Sahelian Zone of Cameroon: The Case of the City of Maroua and Garoua
    
    AU  - Kodji Deli
    AU  - Etienne Tchoffo Houdji
    AU  - Albert Ayang
    AU  - Dieudonne Kidmo Kaoga
    AU  - Noel Djongyang
    Y1  - 2024/07/15
    PY  - 2024
    N1  - https://doi.org/10.11648/j.ijrse.20241302.12
    DO  - 10.11648/j.ijrse.20241302.12
    T2  - International Journal of Sustainable and Green Energy
    JF  - International Journal of Sustainable and Green Energy
    JO  - International Journal of Sustainable and Green Energy
    SP  - 28
    EP  - 42
    PB  - Science Publishing Group
    SN  - 2575-1549
    UR  - https://doi.org/10.11648/j.ijrse.20241302.12
    AB  - The main objective of this study is to compare thirty-five (35) solar radiation models available in the open literature in order to predict monthly solar radiation in two main cities of Cameroon. This estimation and comparison are based on selected statistical comparison parameters named, root mean square error (RMSE), mean bias error (MBE), mean percentage error (MPE) and determination coefficient (R2). These different models are implemented using regression analysis tools named Exel and MATLAB. Estimated values were compared with measured values according to normalized values of statistical parameters, using measured meteorological data of more than 19 years, from 1984 to 2015. All the models have been classified with their associated ranking according to their statistical parameter accuracy. From this study it appears that the models of Ertekin and Yaldiz  (MOD20), Togrul and Onat (MOD28), are more accurate than other models. Indeed, for the city of Maroua (MBE%=-2.82E-14; RMSE%=0.862; MPE=-0.00845; R2=0.985), while for Garoua (MBE%=-9.21E-15; RMSE%=0.806; MPE=-0.00631; R2=0.959). according to their accuracy these models can be therefore be used to predict monthly solar radiation for soudano-sahelian regions of Cameroon. Correlation equations found in this paper will help solar energy researcher to estimate data with trust because of its fine agreement with the observed one. hence the models presented in this study could be used to evaluate accurately the solar radiation at any locations with similar climate.
    
    VL  - 13
    IS  - 2
    ER  - 

    Copy | Download

Author Information