Preliminary results of temperature modelling in Nigeria using neural networks

Okoh, Daniel; Yusuf, Najib; Adedoja, Oluwaseye Samson; Musa, Ibrahim; Rabiu, Babatunde

doi:10.1002/wea.2559

Cited by 10 publications

(9 citation statements)

References 13 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Clearly, the figures do also demonstrate the expectation that neural networks perform better on interpolated predictions than on extrapolated predictions; this is why the RMSEs in Figure are relatively higher than those in Figure . The scenario we explain is rather one in which, including more number of hidden layer neurons (above some limits), causes the networks to more accurately fit interpolated data than it does for extrapolated data as discussed in Okoh et al ().…”

Section: Methods and Resultsmentioning

confidence: 97%

A Hybrid Regression‐Neural Network (HR‐NN) Method for Forecasting the Solar Activity

et al. 2018

View full text Add to dashboard Cite

The Sun is the major driver of space weather events, and as a result, most applications requiring modeling/forecasting of space weather phenomena depend largely on the activities of Sun. Accurate modeling of solar activity parameters like the sunspot number (SSN) is therefore considered significant for the quantitative modeling of space weather phenomena. Sunspot number forecasts are applied in ionospheric models like the International Reference Ionosphere model and in several other projects requiring prediction of space weather phenomena. A method called Hybrid Regression‐Neural Network that combines regression analysis and neural network learning is used for forecasting the SSN. Considering the geomagnetic Ap index during the end of the previous cycle (known as the precursor Ap index) as a reliable measurement, we predict the end of solar cycle 24 to be in March 2020 (±7 months), with monthly SSN 5.4 (±5.5). Using an estimated value of precursor Ap index as 5.6 nT for solar cycle 25, we predict the maximum SSN to be 122.1 (±18.2) in January 2025 (±6 months) and the minimum to be 6.0 (±5.5) in April 2031 (±5 months). We found from the model that on changing the assumed value of precursor Ap index (5.6 nT) by ±1 nT, the predicted peak of solar cycle 25 changes by about 11 sunspots for every 1‐nT change in the assumed precursor Ap index.

show abstract

Section: Methods and Resultsmentioning

confidence: 97%

A Hybrid Regression‐Neural Network (HR‐NN) Method for Forecasting the Solar Activity

et al. 2018

View full text Add to dashboard Cite

show abstract

“…We used one hidden layer in this work since it has been shown that including more than one hidden layer does not lead to much difference in the accuracy of results (Haykin, 1994), and a possible drawback in using multiple hidden layers is that they are more prone to fall in bad local minima (Liu et al, 2007;Okoh, Yusuf, et al, 2015).…”

Section: Training Proceduresmentioning

confidence: 99%

A Neural Network‐Based Ionospheric Model Over Africa From Constellation Observing System for Meteorology, Ionosphere, and Climate and Ground Global Positioning System Observations

et al. 2019

Self Cite

View full text Add to dashboard Cite

The first regional total electron content (TEC) model over the entire African region (known as AfriTEC model) using empirical observations is developed and presented. Artificial neural networks were used to train TEC observations obtained from Global Positioning System receivers, both on ground and onboard the Constellation Observing System for Meteorology, Ionosphere, and Climate satellites for the African region from years 2000 to 2017. The neural network training was implemented using inputs that enabled the networks to learn diurnal variations, seasonal variations, spatial variations, and variations that are connected with the level of solar activity, for quiet geomagnetic conditions (−20 nT ≤ Dst ≤ 20 nT). The effectiveness of three solar activity indices (sunspot number, solar radio flux at 10.7-cm wavelength [F10.7], and solar ultraviolet [UV] flux at 1 AU) for the neural network trainings was tested. The F10.7 and UV were more effective, and the F10.7 was used as it gave the least errors on the validation data set used. Equatorial anomaly simulations show a reduced occurrence during the June solstice season. The distance of separation between the anomaly crests is typically in the range from about 11.5 ± 1.0°to 16.0 ± 1.0°. The separation is observed to widen as solar activity levels increase. During the December solstice, the anomaly region shifts southwards of the equinox locations; in year 2012, the trough shifted by about 1.5°and the southern crest shifted by over 2.5°. Key Points:• The first regional TEC model over the entire African region using empirical observations is developed • The model offers opportunities to conduct high spatial resolution investigations over the African region • EIA occurrence is reduced during the June solstice, and the anomaly region shifts southwards during December solstice Data used in this work include GPS data, indices for solar and geomagnetic activities, and data from ionospheric models used to comparatively verify/validate the model developed. Figure 7. RMSE variations for predictions of the AfriTEC model using the test data set under conditions of varying (a) latitudes, (b) F10.7 values, (c) local times, and (d) days of the years.Figure 11. (a) Sample TEC profile for longitude 20°E illustrating the determination of anomaly crest and trough locations. The illustrated profile is for the March equinox day of year 2012. (b) to (d) are spatial simulations of TEC from the AfriTEC model for 13:00 UT of day number 79 of years 2009, 2012, and 2014, respectively. The F10.7 values are respectively 68, 101, and 150.

show abstract

“…The feedback forward back-propagation neural network training technique [10 , 11] was used. In recent times, previous studies by [12] , [13] , [14] have also used neural networks to train atmospheric dataset collected from the region, and their results indicate that neural networks are good candidates for atmospheric modeling. In particular, [15] posits that the neural network approaches are more accurate than traditional, regression-based approaches.…”

Section: Methodsmentioning

confidence: 99%

“…The Levenberg-Marquardt (LM) back-propagation algorithm [16] was used for training. The algorithm is suitable for the training due to its speed and efficiency in learning [13 , 17 , 18] . References [19] , [20] , [21] contain more general information on neural networks.…”

Section: Methodsmentioning

confidence: 99%

A median absolute deviation-neural network (MAD-NN) method for atmospheric temperature data cleaning

et al. 2021

Self Cite

View full text Add to dashboard Cite

Some of the biggest challenges in climate change arise from bad dataset. To address this issue, we have developed a novel method for cleaning coarse atmospheric dataset; the median absolute deviation-neural network (MAD-NN) method. By combining the median absolute deviation (MAD) technique with neural network training, this method uses a sequence of steps to clean coarse atmospheric dataset and to predict high accuracy dataset for periods when measurements are not available. To demonstrate this method, we used atmospheric temperature data for 17 different observational weather stations across Nigeria. In brief: We developed a novel method for generating consistent data stream from coarse dataset. The MAD-NN method can be used to fill observational data gaps and remove spikes in data. This method is specifically useful for weather observatories with coarse atmospheric data, as well as increasing the credibility of scientific findings.

show abstract

Preliminary results of temperature modelling in Nigeria using neural networks

Cited by 10 publications

References 13 publications

A Hybrid Regression‐Neural Network (HR‐NN) Method for Forecasting the Solar Activity

A Hybrid Regression‐Neural Network (HR‐NN) Method for Forecasting the Solar Activity

A Neural Network‐Based Ionospheric Model Over Africa From Constellation Observing System for Meteorology, Ionosphere, and Climate and Ground Global Positioning System Observations

A median absolute deviation-neural network (MAD-NN) method for atmospheric temperature data cleaning

Contact Info

Product

Resources

About