^{1}

^{*}

^{1}

^{2}

The paper assessed the effect of variability of climatic elements on Agulu lake, Anambra State, Nigeria. Data for the work were acquired from Landsat website (landsat.org) for 35 years. Monthly records of the elements were collected from the synoptic meteorological station at Amawbia in Awka urban area. The penman model was used in the estimation of open water vapour from the lake. The analysis was performed using the output of the classified satellite imagery which was digitized for the entire year and the area of the lake was calculated for all the years of study. Statistical analysis of satellite imagery was further employed to analyse trends and relationships. Result shows that the trend is positive and significant at 95% confidence level. Fluctuations in the value of temperature and other variables were removed by 5-year moving average. The Agulu lake surface area was found to have contracted from 0.6177 km
^{2} in 1978 to 0.3583 km
^{2} in 2013. Recommendations on how to ameliorate the problem were made.

Lakes are major repositories of biodiversity and provide multiple ecosystem services [

A critical and perhaps the dominant global environmental problem in the last three decades is global warming resulting in global climate change [

In addition, lakes and rivers are important links in the hydrological cycle of the earth and studies have noted that such changes in climate as predicted are capable of altering the hydrological and other physical features of lakes [

This, however, necessitated the need for studies to examine how changes in lake size as a result of shifts in water balance due to climatic change are capable of making the lacustrine ecosystem vulnerable to climate change, while these changes are likely to affect the species composition [

Agulu Lake is a natural inland water lake characterized by sacred and cultural landmark of attraction which is gradually being devastated by natural and human factors of flooding, soil/gully erosion as well as landslides due to poorly consolidated geological formations, weathering and pollution [^{2} while its surface area has varied over time. The annual rainfall of the study area ranges from 1383 mm to 2090 mm while the mean annual rainfall is about 1851.9 mm. The mean maximum and minimum temperatures are 32.1˚C and 23.5˚C. Rainfall constitutes the main source of precipitation to the lake. The Nanka sands (the main aquifer of Agulu lake hydrological system) measures 5000 m × 100 m × 30 m [^{7} m^{3} [^{7} m^{3}/yr., an input into the lake. The lake is

the source of Idemili River which drains different communities in Anambra state before emptying into the River Niger.

The Landsat data were acquired from the Landsat.org which comprised of the Multi Spectral Scanner (MSS), thematic mapper (TM), Enhance Thematic Mapper plus (ETM+) image and the Operational land Imager (OLI). The images were acquired to cover 36 years period. The satellite data have 30m spatial resolutions and the TM, ETM and ETM+ images have spectral range of 0.45 - 2.35 μm with bands 1, 2, 3, 4, 5, 6, 7 and 8 while the Operational Land Imager (OLI) extends to band 12. For extracting water surface feature, data from the available bands 2, 3 and 4 of the Landsat imagery of the study area were used due to its higher ability to discriminate water and land area.

Monthly records of air temperature, relative humidity, sunshine duration, wind speed and rainfall were collected from the meteorological archive of the Nigerian meteorological station closest to the lake and estimates of open water evaporation from the lake were made using the mean monthly meteorological data available for 36 years. The penman model was used in this regard.

The temporal variations in the data set for precipitation, temperature and the estimated open water evaporation are investigated using time series analysis while a trend analytical technique was carried out to test the significance of the trend in temperature, open water evaporation and precipitation characteristics over the study area by calculating the linear regression and correlation coefficients of the data sets over time. ANOVA table was used to determine the significance of the regression equation between dependent and independent variables while the assessment indices used in this work such as RMSE, PBIAS, NSE and SE indicate that the model represents the observed data quite well.

Annual mean temperature data for a period of 36 years between 1978 and 2013 shown in

The fluctuations in the value of temperature in the study area such that, in order to easily determine the general trend of these values, we employed a 5 year moving average to smoothen these fluctuations (

Dividing the study period in to three decades, it appears that, in support of the result shown above, temperature is truly increasing. From 1978 to 1987, an increasing trend in temperature was observed such that by 1987, it has exceeded the mean temperature for the decade (27.5˚C) and that of the 36-year period under study. Thus, between 1978 and 1987, temperature increased by 1.3˚C. From 1988 to 1997, there was less fluctuations in temperature value while the mean temperature for this period is 27.9˚C. In 1998, an all-time high temperature (28.9˚C) for the 36-year period was recorded for the study area and also, highest temperature values throughout the period were maintained from 1998 to 2007. The mean temperature for this third decade is 28.24˚C. The overall mean temperature as shown in the trend plot above is 27.9˚C. However, temperature is generally on the increase in the study area and this appears to be in agreement with the study by [

The result shows that there is marked change in the magnitude and seasonal temperature pattern. The temperature pattern shows that temperature, during the months of January-May, October-December, has increased. The months of June, July, August and September have no marked change in temperature.

Evaporation values for the lake were estimated at monthly time step using Penman equation from1978 to 2013. The monthly values of each year were summed for the 36-year period to generate the annual time series data for lake evaporation. Annual fluctuations of the estimated open water evaporation are shown in

We applied a 5-year moving average to produce a more smoothened sequence of the data as shown in

this trend was tested using “t” test and was found not to be statistically significant at 0.05 confidence level. The observed increase in evaporation coupled with the observed change in temperature pattern as depicted in

We performed a correlation analysis to examine the relationship between variations observed in the evaporation data and temperature. The result of the analysis (

Furthermore, monthly mean for evaporation was calculated for the period of study and a graphical interpretation of the variation is shown in

The precipitation data were summed anually to generate annual time series of rainfall from 1978-2013. The trend plot of rainfall achieved is shown in

The high irregular nature of the plot demanded the application of a smoothening function which is the moving average. The plot of a 5-year moving

Evaporation | Evap. | Temp. | CD | N |
---|---|---|---|---|

1.00 | 0.61 | 37% | 36 |

average of rainfall is shown in

Lake Surface Area

Agulu lake surface area data were extracted from Landsat imageries from 1978 to 2013 at interval of 5years. Data were obtained for 1978, 1983, 1988, 1993, 1998, 2003, 2008 and 2013.

By means of linear interpolation, we estimated for missing data on lake surface area for the years for which they are not available. The trend plot of the values is shown in

The statistical results showed that the surface area of the lake was about 0.6177 km^{2} in 1978, 0.3746 km^{2} in 2008 and 0.3583 km^{2} in 2013 (Figures 11-13). The result further showed that the lake surface area changed by 0.0289 km^{2} between 1978 and 1983, 0.1065 km^{2} between 1983 and 1988, 0.0584 km^{2} between 1988 and 1993, 0.0012 km^{2} betwen 1993 and 1998, 0.0084 km^{2} between 1998 and 2003, 0.0397 km^{2} between 2003 and 2008, and 0.0163 km^{2} betweeen 2008 and 2013. The most intense change occurred between 1983 and 1988 during which the lake lost about 0.1065 km^{2}, approximately 17%, of its surface area. On a general note, between 1978 and 2013, the lake lost 42% of its surface area. However, the change in surface area of the lake for the years 1978 to 2013 is shown in

Multiple correlation and regression was performed to examine the extent of relationships between surface area of agulu lake and climatic variables namely: rainfall, evaporation and temperature in the past 36 years. The multiple correlation analysis yielded a correlation coefficient of r = 0.685. This implies that the combined degree or strength of relationship between surface area of Agulu lake and the sets of independent variable of rainfall, temperature and open water evaporation is 0.685. The coefficient of multiple determination, R^{2}, is given as R^{2} = 0.47. In addition, the value of the coefficient of multiple determination revealed that, out of 100% of numerous factors affecting the surface area of Agulu lake, 47% of such variations are explained by the combined variations of temperature, rainfall and evaporation. However, it showed that other factors such as anthropogenic activities [

independent variable (rainfall, temperature and evaporation), a mathematical function was fitted to the sets of variables such that we can use values of the three independent variables to predict values of surface area of the lake. The multple regressive equation developed from this relationship is given below as follows:

Y = 3.619 − 0.000059 X 1 − 0.0902 X 2 − 0.00043 X 3

where, Y = Surface area, X_{1} = Rainfall, X_{2} = Temperature and X_{3} = Evaporation.

The regressive model was further used to predict for surface area for the period, 1978 to 2013.

The significance of the regression equation was tested by means of analysis of variance test. A null hypothesis is posited as: “there is no significant linear relationship between the dependent variable and the independent variables”. At 0.05 level of confidence, the critical value of F from SNEDECOR’s table is 2.90. Since F of 9.436 > 2.90, there is a significant linear relationship between surface area and the three independent variables. There was a good correlation between the predicted values and the original lake area data (r = 0.68) (

Year | Observ. Area (km^{2}) | Predicted est. |
---|---|---|

1978 | 0.618 | 0.53964 |

1988 | 0.482 | 0.428145 |

1998 | 0.423 | 0.336394 |

2008 | 0.375 | 0.50572 |

2013 | 0.358 | 0.464213 |

Model | Sum of Squares | Df | Mean Square | F | Sig. |
---|---|---|---|---|---|

Regression | 0.115 | 3 | 0.038 | 9.436 | 0.000^{b} |

Residual | 0.130 | 32 | 0.004 | ||

Total | 0.245 | 35 |

Evaluating the Performance of the Model

The accuracy measure of this model was evaluated using the Nash-Sutcliffe efficiency (NSE) ratio [

This study has assessed the effects of climate variability and the potential contribution of human activities to the observed variations in the surface area of Agulu Lake in Anambra state using land based station data; a lake currently being converted into a tourist resort. The nature of physical relationship between this morphometric index and selected climatic variables (rainfall, evaporation and temperature) was established. A multiple regression model for the surface area change was derived using recent climate data. It has been shown that climate change effects are not independent of other human stresses on the system, nor are the effects of these stresses independent of climate change effects [

Based on the foregoing, the following recommendations are made:

1) This paper has provided preliminary information on the current state of the lake and its potential response to changes in climate. The model gives a general behaviour of the lake’s surface area to changes in climatic variables. Therefore, for ultimate performance of the model, it is recommended that this model be improved upon by considering other factors not considered in this study.

2) Since the state government is currently harnessing the tourism potentials of the lake, it is imperative that the potential threat posed by climate change on the long term economic viability of this project be acknowledged and concerted efforts be made by the government to address this potential threat. Climate oriented policies bordering on mitigation and adaptation and good conservation principles should be pursued and effort should be made towards provision of improved and efficient water supply scheme in the study area and environs to checkmate and avoid excessive water abstraction from the lake.

3) There is need for government to improve the network density of hydro-meteorological stations within and around inland water bodies scattered across the country for improved gathering of hydro-meteorological information. This will ensure accurate forecasting of the quantity and quality of these water resources. For instance, installation of gauging stations within the lake and along Idemili river will provide reliable information on lake water level and lake discharge respectively; thus, ensuring effective monitoring of changes in the lake round the year and this will lead to improved modeling.

Nzoiwu, C.P., Ezenwaji, E.E. and Okoye, A.C. (2017) A Preliminary Assessment of the Effects of Climate Variability on Agulu Lake, Anambra State, Nigeria. American Journal of Climate Change, 6, 694-710. https://doi.org/10.4236/ajcc.2017.64035