Volume 3, Number 14, December 2022

e-ISSN: 2797-6068 and p-ISSN: 2777-0915

 

 


ANALYSIS OF SATELLITE RAIN DATA USAGE

ON THE RATIONALIZATION ACTIVITIES OF THE RAIN POST NETWORK (CASE STUDY: RATIONALIZATION OF THE JELAI WATERSHED RAIN POST NETWORK)

 

Ari Susanto, Wateno Oetomo, Esti Wulandari

Universitas 17 Agustus 1945 Surabaya, Indonesia

Email: [email protected], [email protected], [email protected]

 

 

KEYWORDS

rainfall, GPM, validation, Kagan

ABSTRACT

An alternative solution to the availability of inadequate rain data as hydrological data input is with the help of Global Precipitation Measurement (GPM) satellite rainfall data using remote sensing technology (satellite). The purpose of this study was to find correlations and corrections of data and validate GPM satellite data with rainfall data at rain stations and observation data in the Jelai watershed. The corrected GP M rain data validation results in Nash-Sutcliffe Efficiency (NSE), Root Mean Squared Error (RMSE), Correlation Coefficient (R), and Relative Error (KR). The validation results resulted in NSE values of 0.33, RMSE 48.54, Correlation Coefficient (R) of 0.75, and Relative Error of 0.19 for 2019 and yielded NSE values of -0.14, RMSE 100.24, Correlation Coefficient (R) of -0.36, and Relative Error of 0.23 for 2020. The overall analysis shows that GPM data can be used as an alternative to rain data if in a watershed there is a small number of rain posts that do not meet the WMO criteria. As a suggestion for further research, it is necessary to calibrate and validate by distinguishing between rain data in wet years and dry years

 

INTRODUCTION

Precipitation data information is very important for various analysis of water resources. Rainfall data can be temporal (time series) or spatial (Renaldhy et al., 2021). As one of the important data in hydrological analysis, rainfall data is obtained from measurements at the rain station post, so that the rainfall data obtained is expected to have sufficient accuracy (Abdaa et al., 2021).

Rainfall data in time series recording can provide trend information from the nature of rain in a place whether it has increased or vice versa (Arrokhman et al., 2021). From this description it can be said that rainfall data is quite important climatological data. Accurate and timely regional and global precipitation observations and forecasts are essential for a wide range of research and applications (Astuti et al., 2022).

In fact, to obtain representative rainfall observation data that is both in terms of quality and quantity or duration of observation data that is sufficient according to the requirements is very difficult (Azka et al., 2018). It is difficult to obtain rainfall data, due to the limited number of measuring instruments or gauges, especially in remote areas, so it will be difficult to conduct a study and analysis of water resources based on rainfall data in one place because not all places have manual or automatic rainfall monitoring stations (Derin et al., 2016).

According to Syaifullah, the latest technological developments, namely in the form of satellite technology (remote sensing) are able to make breakthroughs in terms of obtaining rainfall information (precipitation) because the current remote sensing technology has been able to measure rainfall from a distance (Faisol & Bachri, 2021). Areas that do not have adequate rain recording stations are almost impossible to measure rainfall, but with this technology it is possible to obtain rainfall data that is not limited in space and time, so that in simple terms it can be said that with satellite technology rainfall data can be obtained at any time. anywhere and anytime (Oktaverina et al., 2022).

On February 27, 2014, NASA and JAXA launched Global Precipitation Measurement (GPM) as a successor to the TRMM satellite (Sarwanta & Abdulgani, 2021). The aim of this satellite launch is to improve the quality of precipitation observations on a global scale. GPM has a global coverage of 65o North Latitude to 65o South Latitude with observations every 3 hours (Pangestu et al., 2021). The GPM constellation can estimate the intensity and type of precipitation, cloud structure in 3 dimensions, storm systems, microscopic ice and liquid in clouds, and the amount of precipitation that falls on the earth's surface (Orfa & Samad, 2019).

However, before the GPM satellite rain data can be used, it is necessary to evaluate whether the rainfall data from the GPM satellite and from the existing rain station post network will produce maximum information so that the amount of rainfall can be obtained at all points with sufficient accuracy or even very different (Tang et al., 2020). �In the Jelai River Basin with an area of 7,682 km2 there are three rain stations which are within the DAS (Prabawadhani et al., 2016).

This study will examine how the correlation of postal rainfall station data with satellite rainfall data. After obtaining the corrected GPM satellite data, it will then be used to make an annual rainfall isohyet map with the location/network of rainfall posts based on the Kagan Method (Suryaningtyas, 2019).

 

RESEARCH METHOD

Consistency Test

A data consistency test was carried out to find out whether there are deviations in the available rainfall data, so that it can be seen whether the data is suitable for use in further hydrological analysis or not. In this study, 2 (two) methods were carried out, namely (1) multiple mass curves; (2) Rescaled Adjusted Partial Sums (RAPS).

Homogeneity Test

A series of hydrological data presented chronologically as a function at the same time is called a periodic series. The data is arranged in a series of periodic forms, so that it must be tested before being used for further analysis. The intended data tests are: (1) Test for the Absence of Trend; (2) Stationary Test; (3) Persistence Test. The three stages of testing are often referred to as data filtering.

GPM Rainfall Data Validation Test

For validation tests, the Nash-Sutcliffe Efficiency (NSE), Correlation coefficient (R), Root Mean Squared Error (RMSE) and Relative Error (RE) methods are used. Two validation analyzes were carried out, namely validation of GPM data that had not been corrected and validation of GPM data that had been corrected.

Uncorrected GPM data validation using uncorrected GPM and rain station rainfall data. The period used is monthly with a data length of 11 years.

As for the validation of the corrected TRMM data, a number of processes are carried out first, namely calibration, verification and validation. Calibration and verification using the scatter plot method. For calibration, a monthly period is used with a data length of 11 years (2007 and 2009-2018). While the verification and validation tests use a monthly period with a data length of 2 years (2019-2020) excluding the calibration year.

The validation method formula used in this study is:

1) ������� Nash-Sutcliffe Efficiency (NSE)

This method shows how well it plots the observed (measurement) values compared to the simulated-predicted values, according to the 1:1 line, with a range of values ∞ to 1. In other words, the closer to 1, the better the NSE value.

With:

Xi ������ = observation data (actual data)

Yi ������ = estimated data (estimated yield data)

Xi ������ = average observation data

N ������� = amount of data

 

Table 1

Nash-Sutcliffe Efficiency (NSE) Score Criteria

2) ������� Correlation Coefficient

The purpose of this analysis is to obtain patterns and close relationships between two or more variables.

With:

Xi ������ = observation data (actual data)

Yi ������ = estimated data (estimated yield data)

N ������� = amount of data

Table 2.2 Correlation Coefficient Value Criteria

3) ������� Root Mean Squared Error (RMSE)

With:

Xi ������ = observation data (actual data)

Yi ������ = estimated data (estimated yield data)

N ������� = amount of data

4) ������� Relative Error Test

This test is used to determine the comparison between the magnitude of one variable to another variable that is used as a benchmark for the actual variable.

With:

Xi ������ = observation data (actual data)

Yi ������ = estimated data (estimated yield data)

N ������� = amount of data

Thiessen Polygon Method

This method is used to calculate the area's average rainfall, where in a watershed there are several rain posts.

Kagan's method

With the Kagan method, the ideal distance between the locations of the automatic rain posts and the distribution of the locations of the automatic rain posts can be identified.

Isohyet Map

After the distribution of rainfall stations is known based on the Kagan Method, an Isohyet Map will be made in the Jelai watershed based on the rainfall data and the corrected GPM satellite rain data, so that comparisons/differences can be identified.

 

 

 

 

 

 

 

 

 

 

RESULTS AND DISCUSSION

Hydrological Analysis

Consistency Test

The consistency test was carried out using two methods, the multiple mass curve method and the RAPS method.

�

Figure 1

The Double Mass Curve of the Eye Sweet Rain Post, the Nibung Island Rain Post and the Pasir River Rain Post

 

Table 2

Recapitulation of α values at each rain station post

No

Rain Station Post

Marka

R value 2

1

Sweet Eyes

44.46 o

0.9335

2

Nibung Island

48.70o _

0.9889

3

Sand River

43.89o _

0.9295

Source: Analysis results, 2022

 

Figure 2

�Grid-48, Grid-78 and Grid-90 Multiple Mass Curves

 

Table Error! No text of specified style in document.

Recapitulation of α values in each gris (GPM data)

No

Grid Number

Marka

R value 2

1

Grid-48

47.32o _

0.9990

2

Grid-78

44.39 o

0.9988

3

Grid-90

43.22o _

0.9996

Source: Analysis results, 2022

Information:

Grid-48 �GPM grid that corresponds to the location of the Sweet Rain Eye Post

Grid-78 �GPM grid corresponding to the location of the Nibung Island Rain Post

Grid-90 �grid GPM which corresponds to the location of the Pasir Sungai Rain Post

 

 

 

 

 

 

 

 

Table 4

Recapitulation of Consistency Test Results

No

Name

Curve Method

RAPS method

Ket.

Post

Double Mass

 

Corner

Q/n 0.5 count

Q/n 0.5 table

R/n 0.5 count

R/n 0.5 table

1

Sweet Eyes

44,46

0.413

1.172

0.778

1,340

Consistent

2

Nibung Island

48,70

0.731

1.116

1,080

1.235

Consistent

3

Sand River

43.89

0.505

1.116

0.947

1.235

Consistent

4

Grid-48

47,32

0.577

1.172

0.888

1,340

Consistent

5

Grid-78

44,39

0.412

1.172

0.787

1,340

Consistent

6

Grid-90

43,22

0.628

1.172

0.993

1,340

Consistent

Source: Analysis results, 2022

Based on Figure 1 and Figure 2 as well as Table 1 and Table 2, it can be said that the post rainfall data of the rain station and the GPM data used after being tested using the Multiple Mass Curve Method are consistent because the resulting angles are in the range of values 42o < α < 48o . Meanwhile, based on Table 3, the rainfall data consistency test using the RAPS method also meets the test requirements because the Q count <Q table and R count <R table so that the results can be considered consistent.

The results of this test indicate that the selected data can be used for further hydrological testing and analysis.

 

Homogeneity Test

In this study, the annual rainfall data of rainfall stations were tested for no trend using the Spearman method using a 2-tailed T-Test. Recapitulation of test results is presented as follows.

Table 5

Summary of Absence Test Results for Annual Period

No

Post Name

t count

a

t c

Information

1

Sweet Eyes

-1,582

5%

2,179

Does not show a trend

2

Nibung Island

-2,840

5%

2,571

Does not show a trend

3

Sand River

-1.419

5%

2,571

Does not show a trend

4

GPM Grid-48

0.206

5%

2,179

Does not show a trend

5

GPM Grid-78

0.175

5%

2,179

Does not show a trend

6

GPM Grid-90

0.659

5%

2,179

Does not show a trend

Source: Analysis results, 2022

Based on Table 5 it can be seen that all data does not show a trend by showing t count <t table at the 5% confidence level. Thus, the data can be analyzed further.

 

Table 6

Summary of Variance Stability Test Results (Test F) Annual Period

No

Post Name

F count

a

Fc _

Information

1

Sweet Eyes

1.013

5%

4,280

The variance value is stable

2

Nibung Island

0.095

5%

19,160

The variance value is stable

3

Sand River

0.496

5%

19,160

The variance value is stable

4

Grid-48

1,638

5%

3,410

The variance value is stable

5

Grid-78

0.743

5%

3,410

The variance value is stable

6

Grid-90

1,339

5%

3,410

The variance value is stable

Source: Analysis results, 2022

 

Table 7

Summary of Average Stability Test Results (t test) Annual Period

No

Post Name

t count

a

t c

Information

1

Sweet Eyes

-0.400

5%

2,179

The average value is stable

2

Nibung Island

-1,713

5%

2,571

The average value is stable

3

Sand River

-2,407

5%

2,571

The average value is stable

4

Grid-48

0.216

5%

2,179

The average value is stable

5

Grid-78

0.328

5%

2,179

The average value is stable

6

Grid-90

0.363

5%

2,179

The average value is stable

Source: Analysis results, 2022

 

From Table 6 and Table 7 above it can be seen that the calculated F value < F table value and the t calculated value < t table value , so it can be concluded that the rainfall data from the three rain station posts and the GPM rainfall data used have variance and average stable average. The persistence test is an independent test for each value in the periodic series. First, the number of serial correlation coefficients must be calculated using the Spearman method, then the persistence test is calculated using the T-Test. Recapitulation of test results is presented as follows.

 

Table 8

Summary of Annual Period Persistence Test Results

No

Post Name

t count

a

t c

Information

1

Sweet Eyes

0.492

5%

2,179

Data is random

2

Nibung Island

1.116

5%

2,571

Data is random

3

Sand River

1.007

5%

2,571

Data is random

4

Grid-48

-0.038

5%

2,179

Data is random

5

Grid-78

-0.437

5%

2,179

Data is random

6

Grid-90

0.099

5%

2,179

Data is random

Source: Analysis results, 2022

 

Table 8 it can be seen that almost all of the data is random by showing tcount <ttable at the 5% level of confidence. Thus, the data can be analyzed further.

 

Rain Station Data Correlation Rain Station and GPM

Based on Table 8, the results of the correlation analysis of all rain posts with GPM data have a good correlation for the Sweet Eyes Rain Post and the Nibung Island Rain Post (the correlation coefficient value is at a value > 0.6), while for the Sungai Pasir Rain Post it is not good ( the value of the correlation coefficient which is at a value of <0.6).

 

Table 9

Recapitulation of Correlation Results for Annual, Monthly, and Monthly Average Data

Source: Analysis results, 2022

 

 

 

GPM Rain Data Calibration

Rainfall calibration uses an 11-year period (2007 and 2009-2018 rain data) from the Manis Mata rain post, while for verification and validation it uses 2019 and 2020 rain data. This is determined by considering the length of the data and the correlation results in Table 9

Table 9 and Table 10 show the results of the regression equation and the resulting coefficient of determination (R 2 ). From the regression equation that has been obtained to obtain the corrected GPM rain data, then the regression equation with the largest R value is used. The results of the GPM rainfall regression equation in the Jelai watershed with R� = 0.6120 with the intercept linear equation (when using monthly rainfall data) and R 2 = 0.9788 with the intercept linear equation (when using monthly average rainfall data). Because the value of R2 with monthly average data is greater than using monthly data, then to correct the GPM rain data use the equation: y = 0.4762x.

 

Table 10

Tabulation of GPM Monthly Rainfall Regression Equation Results in Jelai Watershed

No

Regression Equation

�y� value

Value �R 2 �

1

linear

y = 0.3796x + 27.555

0.1625

2

Linear Intercepts

y = 0.4684x

0.6120

3

Logarithmic

y = 65.631ln(x) � 228.26

0.1609

4

Polynomial

y = -0.0007x 2 + 0.744x - 9.1964

0.1730

5

Polynomial Intercept

y = -0.0006x2 + 0.6746x

0.1728

6

rank

-

-

Source: Analysis results, 2022

 

Table 11

Tabulation of GPM Monthly Mean Rainfall Regression Equation Results in Jelai Watershed

No

Regression Equation

�y� value

Value �R 2 �

1

linear

y = 0.3464x + 35.98

0.7580

2

Linear Intercepts

y = 0.4762x

0.9788

3

Logarithmic

y = 73.743ln(x) � 280.32

0.6869

4

Polynomial

y = 0.0012x 2 � 0.2393x + 99.385

0.8025

5

Polynomial Intercept

y = -0.0005x2 + 0.6175x

0.7214

6

rank

y = 4.0806x 0.6167

0.7341

Source: Analysis results, 2022

 

 

 

GPM Rain Data Verification

For verification, rain data for 2019 and 2020 was used. The following is a graph of rain data for 2019 and 2020 for the Sweet Rain Post and the corrected GPM data (grid-48).

 

 

 

 

 

 

 

 

 

 

 


Figure Error! No text of specified style in document.

Graph of Corrected GPM Rainfall in 2019 and 2020

 

Source: Analysis results, 2022

Figure 3

��GPM Rainfall Verification for 2019 and 2020

 

Based on Figure 3 and Figure 4 it can be seen that for 2019 it produces a greater correlation value than in 2020.

 

GPM Rain Data Validation

Validation is carried out on data outside of the data used for calibration (2019 and 2020). To be able to measure the magnitude of the difference in the results of corrected GPM rainfall calculations against postal rainfall data, a mathematical model validation test can be used using Nash Sutcliffe Efficiency (NSE), Root Mean Square Error (RMSE), Correlation Coefficient, and Relative Error. The smaller the RMSE value, the closer the simulated data is to the observed data, conversely the greater the NSE value (maximum equal to 1), the closer the simulation results to the observations.

 

Table 12

GPM Data Validation Test Recapitulation Before and After Correction

2019 year

2020 year

 

Before

After

 

Before

After

NSE

-3.61

0.33

NSE

-2.31

-0.14

RMSE

127,26

48,54

RMSE

170.38

100.24

KR

-0.70

0.19

KR

-0.61

0.23

R

0.75

0.75

R

-0.36

-0.36

Source: Analysis results, 2022

 

Based on the results of Table 12, it is known that the results of the corrected GPM data are better than before being corrected. The validation results in 2019 were better than in 2020. Thus the other grid GPM data will be corrected using the equation y = 0.4762x.

 

Determination of Rainfall Post Network Kagan Method

Based on several ways of determining existing rainfall postal networks, Kagan's method is relatively simple, both in terms of understanding and calculation procedures. Besides being able to produce the required number of posts with a certain level of accuracy, the Kagan method can also provide a clear pattern of placement of rainfall posts. Based on the WMO criteria, it is known that with the condition of the plains and the area of the Jelai watershed 7,682 km 2 , a minimum of 9 automatic rain posts and a maximum of 13 automatic rain posts are needed. Using the Kagan formula (l = ), the distance between posts is 32.10 km (if there are 9 postal units) and 26.21 km (if there are 13 postal units).

 

Source: Analysis results, 2022

Figure 4

�Kagan Triangle For Number of Rain Post 9 Units

Source: Analysis results, 2022

Figure 5

��Kagan Triangle For Rain Post Number 13 Units

Regional Rainfall Calculation

Regional average rainfall analysis or regional rainfall analysis in this study was carried out on postal rain data and corrected GPM data. Regional rainfall analysis in this study uses the Thiessen polygon method, which in principle is to create an area of influence for each rain station post on the watershed area under review.

Source: Analysis results, 2022

Figure 6

��Thiessen Polygon Existing Rain Post

Source: Analysis results, 2022

Figure 7

��Thiessen Polygon 9 Recommended Rain Post

 

Source: Analysis results, 2022

Figure 8

��Thiessen Polygon 13 Recommended Rain Posts

 

 

 

Table 13

Rain Region With Rain Post Data

Source: Analysis results, 2022

 

Table 14

Regional Rain with GPM Data 9 Kagan Recommendation Posts

Source: Analysis results, 2022

 

Table 15

Regional Rain with GPM Data 13 Kagan Recommendation Posts

Source: Analysis results, 2022

 

Based on Tables 13 to 15 it can be seen that the rainfall in areas with postal rain data is greater than using corrected GPM rain data.

Isohyet Map Creation

Isohyet is a line on the map to connect positions that have the same rainfall value. The following isohyet map based on rain post data and corrected GPM rain data (with rain post locations according to Kagan's recommendations).

Source: Analysis results, 2022

Figure 10

��Map of Annual Average Rainfall Isohyet Rain Post Data

 

Source: Analysis results, 2022

 

Figure 9

��Map of Annual Average Rainfall Isohyet GPM Data 9 Pos

 

Source: Analysis results, 2022

 

CONCLUSION

The results of the correlation analysis of GPM satellite rainfall data and rainfall data from the rain station post have good results when using annual rainfall data and monthly average rainfall data. The results of the validation of rainfall data for the Manis Mata station post with the GPM show that the results of the corrected data validation have better results than the GPM data before being corrected. The validation for 2019 is better than for 2020. This shows that the validation of rain data does not always produce good results for all validation years, and further research is needed regarding the validation of rain data which are included in the category of wet years and dry years.

 

REFERENCES

 

Abdaa, D., Fauzi, M., & Sandhyavitri, A. (2021). Rasionalisasi Kerapatan Stasiun Hujan Wilayah Sungai Rokan Berdasarkan Data Hujan Harian dan Variasi Tingkat Kesalahan. JURNAL SAINTIS, 21(02), 61�70.

Arrokhman, N. A., Wahyuni, S., & Suhartanto, E. (2021). Evaluasi Kesesuaian Data Satelit untuk Curah Hujan dan Evaporasi Terhadap Data Pengukuran di Kawasan Waduk Sutami. Jurnal Teknologi Dan Rekayasa Sumber Daya Air, 1(2), 904�916.

Astuti, I. W., Suhartanto, E., & Fidari, J. S. (2022). Rasionalisasi Jaringan Pos Hujan Dan Pos Duga Air Dengan Metode Stepwise Dan Standar WMO (World Metrological Organization) Di DAS Opak. Jurnal Teknologi Dan Rekayasa Sumber Daya Air, 2(1), 377�385.

Azka, M. A., Sugianto, P. A., Silitonga, A. K., & Nugraheni, I. R. (2018). Uji akurasi produk estimasi curah hujan Satelit GPM IMERG di Surabaya, Indonesia. Jurnal Sains & Teknologi Modifikasi Cuaca, 19(2), 83�88.

Derin, Y., Anagnostou, E., Berne, A., Borga, M., Boudevillain, B., Buytaert, W., Chang, C.-H., Delrieu, G., Hong, Y., & Hsu, Y. C. (2016). Multiregional satellite precipitation products evaluation over complex terrain. Journal of Hydrometeorology, 17(6), 1817�1836.

Faisol, A., & Bachri, S. (2021). Komparasi dan Evaluasi Data Hujan Berbasis Satelit dalam Mengestimasi Curah Hujan Harian Maksimum di Provinsi Papua Barat. Prosiding Seminar Nasional Fakultas Pertanian UNS, 5(1), 534�543.

Oktaverina, D. A. R., Suhartanto, E., & Wahyuni, S. (2022). Validasi Data Curah Hujan Satelit TRMM (Tropical Rainfall Measuring Mission) dengan Pos Stasiun Hujan pada Sub DAS Keduang Kabupaten Wonogiri, Provinsi Jawa Tengah. Jurnal Teknologi Dan Rekayasa Sumber Daya Air, 2(1), 265�276.

Orfa, L. E., & Samad, A. (2019). Kajian Penggunaan Data Trmm (Tropical Rainfall Measuring Missions) Untuk Analisis Hidrologi (Studi Kasus pada Bendungan Lolak Kab. Bolaang Mongondow, Sulawesi Utara). Prosiding SENTRA (Seminar Teknologi Dan Rekayasa), 4, 34�39.

Pangestu, I. T., Zakaria, A., & Tugiono, S. (2021). Analisis Korelasi Data Curah Hujan BMKG dengan TRMM (Studi Kasus Stasiun BMKG di Sumatera Utara). Jurnal Rekayasa Sipil Dan Desain, 8(1), 157�180.

Prabawadhani, D. R., Harsoyo, B., Seto, T. H., & Prayoga, B. R. (2016). Karakteristik temporal dan spasial curah hujan penyebab banjir di wilayah dki jakarta dan sekitarnya. Jurnal Sains & Teknologi Modifikasi Cuaca, 17(1), 21�25.

Renaldhy, R., Yasa, I. W., & Setiawan, E. (2021). Evaluasi Rasionalisasi Stasiun Hujan Metode Kagan Rodda dengan Mempertimbangkan Kriteria Penentuan Lokasi Pembangunan Stasiun Hujan. Jurnal Teknik Pengairan: Journal of Water Resources Engineering, 12(1), 49�60.

Sarwanta, S., & Abdulgani, H. (2021). Studi Rasionalisasi Pos Hidrologi Di Wilayah Sungai Noelmina Provinsi Nusa Tenggara Timur. Jurnal Rekayasa Infrastruktur, 7(1), 16�22.

Suryaningtyas, L. (2019). Analisa Hidrologi Data TRMM (Tropical Rainfall Measuring Mission), Di Sub DAS Lesti. Universitas Brawijaya.

Tang, S., Li, R., He, J., Wang, H., Fan, X., & Yao, S. (2020). Comparative evaluation of the GPM IMERG early, late, and final hourly precipitation products using the CMPA data over Sichuan Basin of China. Water, 12(2), 554.

 

 

 

 

 

Copyright holders:

Ari Susanto, Wateno Oetomo, Esti Wulandari (2022)

 

First publication right:

Devotion - Journal of Research and Community Service

This article is licensed under a Creative Commons Attribution-ShareAlike 4.0 International