Peramalan Rata-Rata Suhu Maksimum Bulanan pada Kawasan Urban Heat Island Menggunakan Model GSTARX di Surabaya, Sidoarjo, dan Gresik

Putri, Davina Shafa Felisa (2026) Peramalan Rata-Rata Suhu Maksimum Bulanan pada Kawasan Urban Heat Island Menggunakan Model GSTARX di Surabaya, Sidoarjo, dan Gresik. Other thesis, Institut Teknologi Sepuluh Nopember.

Text 5002221142-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only Download (11MB) | Request a copy

Abstract

Fenomena Urban Heat Island (UHI) menyebabkan suhu di kawasan perkotaan lebih tinggi dibandingkan wilayah sekitarnya akibat penggunaan lahan yang didominasi oleh permukaan keras seperti aspal dan beton. Surabaya, Sidoarjo, dan Gresik merupakan bagian dari kawasan metropolitan di Jawa Timur yang saling berdekatan sehingga perubahan suhu pada satu wilayah dapat berkaitan dengan perubahan suhu di wilayah lainnya. Penelitian ini bertujuan untuk memodelkan dan meramalkan rata-rata suhu maksimum bulanan di ketiga wilayah tersebut menggunakan model Generalized Space-Time Autoregressive with Exogenous Variables (GSTARX). Data yang digunakan merupakan rata-rata suhu maksimum bulanan periode Januari 2016 hingga Februari 2026 yang diperoleh dari ECMWF, dengan 106 bulan sebagai data in-sample dan 16 bulan sebagai data out-sample. Model yang diperoleh adalah GSTARX(3;1) dengan bobot normalisasi korelasi silang yang dihitung dari residual model GSTAR(3;1). Penanganan outlier yang berkaitan dengan fenomena La Niña dan El Niño dilakukan dengan menambahkan variabel dummy intervensi bertipe Additive Outlier (AO) dengan pulse function. Residual model memenuhi asumsi normalitas multivariat dengan proporsi di2 ≤ χ20,5;3 sebesar 54,90%. Nilai MAPE out-sample untuk Surabaya, Sidoarjo, dan Gresik masing-masing sebesar 2,0928%, 2,3309%, dan 1,0922%, sehingga model tergolong sangat akurat. Peramalan periode Maret 2026 hingga Februari 2027 menghasilkan rentang rata-rata suhu maksimum bulanan sebesar 27,2885–27,5921°C di Surabaya, 26,3690–26,8385°C di Sidoarjo, dan 27,4603–27,7182°C di Gresik dengan pola yang relatif stabil antarbulan. Bobot spasial yang terbentuk menunjukkan bahwa pengaruh Sidoarjo terhadap Surabaya dan Gresik lebih dominan dibandingkan pasangan wilayah lainnya. Temuan tersebut sejalan dengan karakteristik fenomena UHI pada kawasan metropolitan yang saling terhubung secara spasial.
==============================================================================================================================
The Urban Heat Island (UHI) phenomenon causes urban temperatures to be higher than surrounding areas due to land use dominated by hard surfaces such as asphalt and concrete. Surabaya, Sidoarjo, and Gresik are part of a metropolitan area in East Java that are located in close proximity, so that temperature changes in one area may be related to temperature changes in other areas. This study aims to model and forecast the monthly average maximum temperature in these three areas using the Generalized Space-Time Autoregressive with Exogenous Variables (GSTARX) model. The data used are monthly average maximum temperatures for the period January 2016 to February 2026 obtained from ECMWF, with 106 months as in-sample data and 16 months as out-sample data. The model obtained is GSTARX(3;1) with normalized cross-correlation weights calculated from the residuals of the GSTAR(3;1) model. Outliers associated with La Niña and El Niño phenomena were handled by adding dummy intervention variables of the Additive Outlier (AO) type with a pulse function. The model residuals satisfy the multivariate normality assumption with a proportion of di2 ≤ χ20.5;3 of 54.90%. The out-sample MAPE values for Surabaya, Sidoarjo, and Gresik are 2.0928%, 2.3309%, and 1.0922%, respectively, indicating that the model is highly accurate. The forecast for the period March 2026 to February 2027 yields monthly average maximum temperature ranges of 27.2885–27.5921°C in Surabaya, 26.3690–26.8385°C in Sidoarjo, and 27.4603–27.7182°C in Gresik with a relatively stable pattern across months. The spatial weights obtained indicate that the influence of Sidoarjo on both Surabaya and Gresik is more dominant compared to other pairs of areas, which is consistent with the characteristics of the UHI phenomenon in spatially connected metropolitan areas.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	Urban Heat Island, GSTARX, suhu maksimum, peramalan spasial temporal, variabel dummy intervensi, Urban Heat Island, GSTARX, maximum temperature, spatio-temporal forecasting, dummy intervention variable
Subjects:	Q Science > QA Mathematics > QA276 Mathematical statistics. Time-series analysis. Failure time data analysis. Survival analysis (Biometry) Q Science > QA Mathematics > QA278.2 Regression Analysis. Logistic regression Q Science > QA Mathematics > QA280 Box-Jenkins forecasting
Divisions:	Faculty of Science and Data Analytics (SCIENTICS) > Mathematics > 44201-(S1) Undergraduate Thesis
Depositing User:	Davina Shafa Felisa Putri
Date Deposited:	24 Jul 2026 03:58
Last Modified:	24 Jul 2026 03:58
URI:	http://repository.its.ac.id/id/eprint/136787

Actions (login required)

View Item