Skip to main content

Official Journal of the Asia Oceania Geosciences Society (AOGS)

Table 1 Examples of univariate time series forecasting in geoscience

From: One-step ahead forecasting of geophysical processes within a purely statistical framework

s/n

Study

Process

Number of original time series

Forecast time scale

Forecast horizon(s) [step(s) ahead]

Univariate time series forecasting method(s)

1

Hong (2008)

Precipitation

9

Hourly

1

(1) Support vector machines

(2) Hybrid model, i.e. a combination of recurrent neural networks and support vector machines

2

Chau and Wu (2010)

2

Daily

1, 2, 3

(1) Neural networks

(2) Hybrid model, i.e. a combination of neural networks and support vector machines

3

Htike and Khalifa (2010)

1

Monthly, biannually, quarterly, yearly

1

Neural networks

4

Wu et al. (2010)

4

Monthly, daily

1, 2, 3

(1) Linear regression

(2) k-nearest neighbours

(3) Neural networks

(4) Hybrid model, i.e. a combination of neural networks

5

Narayanan et al. (2013)

6

Yearly

21 × 3 (months)

AutoRegressive Integrated Moving Average (ARIMA)

6

Wang et al. (2013)

1

Monthly

12

Seasonal AutoRegressive Integrated Moving Average (SARIMA)

7

Babu and Reddy (2012)

Temperature

1

Yearly

10

(1) ARIMA

(2) Wavelet-based ARIMA

8

Chawsheen and Broom (2017)

1

Monthly

121

SARIMA

9

Lambrakis et al. (2000)

Streamflow or river discharge

1

Daily

1

(1) Farmer’s model

(2) Neural networks

10

Ballini et al. (2001)

1

Monthly

1, 3, 6, 12

(1) AutoRegressive Moving Average (ARMA)

(2) Neural networks

(3) Neurofuzzy networks

11

Yu et al. (2004)

2

Daily

1

(1) Support vector machines coupled with an evolutionary algorithm

(2) Standard chaos technique

(3) Naïve

(4) Inverse approach

(5) ARIMA

12

Komorník et al. (2006)

7

Monthly

1, 3, 6, 12

(1) Threshold AutoRegressive (AR) with aggregation operators

(2) Logistic smooth transition AR

(3) Self-exciting threshold AR

(4) Naïve

13

Yu and Liong (2007)

2

Daily

1

(1) Support vector machines coupled with decomposition

(2) Standard chaos technique

(3) Naïve

(4) Inverse approach

(5) ARIMA

14

Koutsoyiannis et al. (2008)

1 × 12 (months) 

Yearly

1

(1) Stochastic

(2) Analogue method

(3) Neural networks

15

Wang et al. (2015)

3

Monthly

12

SARIMA