Rainfall Correction Factor of Chirps Satellite Data Against Observation Data of Ciliwing Watershed (Case Study of Kemayoran Meteorological Statio.
Kurniyaningrum.
Faluty.
Mulya.
Andayani.
Hidayat.
Sejati.
Satar p-ISSN 2580-7552.
e-ISSN 2548-7515.
Volume 9.
Number 2, pp 149 Ae 158, 2024 https://e-journal.
id/index.
php/livas/index FACTOR FOR CORRECTING THE RAINFALL OF CHIRPS SATELLITE DATA
AGAINST OBSERVATION DATA ON THE CILIWUNG WATERSHED
(CASE STUDY OF KEMAYORAN METEOROLOGI STATION)
Endah Kurniyaningrum1*.
Mutiara Difa Faluty2.
Hegi Daniel Mulya3.
Sih Andajani4.
Dina Paramitha Anggraeni Hidayat5.
Wahyu Sejati6.
Hira Sattar7
1,2,3,4,5,6
Department of Civil Engineering.
Faculty of Civil Engineering and Planning.
Universitas Trisakti.
Jakarta, 11440.
Indonesia.
Tokyo Institute of Technology.
Tokyo.
Japan *Corresponding author: kurnianingrum@trisakti.
ABSTRACT
MANUSCRIPT
HISTORY
The hydrological and environmental cycles in a river area strongly affect rainfall intensity and seasonal patterns.
To accurately assess water resource A capacity, precise rainfall data from each observation station is crucial.
However, unevenly distributed rain gauges often challenge researchers, as A insufficient data can hinder their analysis.
In these situations, satellite images A can provide valuable additional information.
Aims: The objective of this study was to analyze the accuracy of CHIRPS satellite rainfall data from observation stations in the Ciliwung watershed, especially in the DKI Jakarta Province area, over the last 30 years .
3Ae2.
Methodology and results: Statistical analysis such as multiple linear regression with the stepwise method is used to analyze CHIRPS rainfall against observed rainfall data according to the location of the rain station.
The validation results in this study show that the average results of the two observation stations have a value of R2 = 0.
91 and NSE = 0.
Conclusion, significance and impact study: CHIRPS data can be categorized as very good if used as an alternative to limited observational rainfall data, which can then be used in analyzing water availability in the Ciliwung watershed (Jakart.
Received May 15, 2024
Revised May 29, 2024
Accepted June 11, 2024
KEYWORDS
Accuracy.
CHIRPS.
Ciliwung Watershed.
Rainfall.
Jakarta
INTRODUCTION
Climate significantly affects many sectors and lives.
It's important to study because climate conditions vary greatly depending on location, latitude, longitude, and the earth's uneven surface .
Rainfall, a key climate component, plays a crucial role.
Changes in rainfall intensity Doi: https://doi.
org/10.
25105/livas.
Rainfall Correction Factor of Chirps Satellite Data Against Observation Data of Ciliwing Watershed (Case Study of Kemayoran Meteorological Statio.
Kurniyaningrum.
Faluty.
Mulya.
Andayani.
Hidayat.
Sejati.
Satar p-ISSN 2580-7552.
e-ISSN 2548-7515.
Volume 9.
Number 2, pp 149 Ae 158, 2024 over 10Ae30 years can impact water availability and human activities .
The world's climate is experiencing changes.
these changes result in increasing temperatures and sea levels.
Future climate change may cause more intensive hydrological cycle processes .
, including increased variations in rainfall .
and changes in evaporation rates .
Increasing global temperatures will increase the rate of evapotranspiration and speed up the water cycle .
As a result, an uneven distribution of moisture in the atmosphere will occur, causing heavy rainfall in one region and extreme drought in other regions .
Changes in rainfall patterns in Indonesia will lead to a delay in the start of the rainy season and a tendency towards an earlier start at the at the end of the rainy season.
This means that the rainy season occurs in a shorter time but has a higher rainfall intensity .
Rainfall is an important climate element for human activities.
Rainfall has characteristics that vary according to space and time, so the availability of adequate data is important for understanding the characteristics of rainfall in a region .
However, the problem is that in some areas, tools for measuring rainfall data are sometimes not available, so to overcome the lack of availability of rain data in recent years, a number of studies have been carried out on the use of rain data based on remote sensing or satellite technology .
Many studies have been conducted regarding the use of satellite rainfall data, including .
, .
, using GPM to measure extreme rainfall events in China and North China .
, using GPM, and PERSIANN to estimate flood discharge in the Progo watershed.
and evaluated satellite rainfall predictions using GPM and PERSIANN .
Rainfall variability is spatially based on the location or place where the rain falls.
mountainous areas, topography, and elevation is a factor that influences bulk rain and should be considered for prediction and mapping.
For the most part, areas have been found that increase rainfall with elevation has a linear relationship.
Apart from the elevation factor, other additional factors that Influencing rainfall are also necessary considered.
This is due to the nature of rainfall in mountainous orographic areas, where the rainfall pattern is complex and incomprehensible with good results .
Mapping the spatial variability of rainfall can be carried out using the method of spatial interpolation.
The resolution of topography has the significant effect on the flood simulation results .
Rainfall is defined as the amount of water that falls to the ground over a specified period of time.
It is typically measured in units of height .
above a horizontal surface, assuming no evaporation, runoff, or infiltration.
The type of vegetation cover can influence rainfall conditions, and precipitation can result in water loss .
and affect flow velocity .
, .
The research has Doi: https://doi.
org/10.
25105/livas.
Rainfall Correction Factor of Chirps Satellite Data Against Observation Data of Ciliwing Watershed (Case Study of Kemayoran Meteorological Statio.
Kurniyaningrum.
Faluty.
Mulya.
Andayani.
Hidayat.
Sejati.
Satar p-ISSN 2580-7552.
e-ISSN 2548-7515.
Volume 9.
Number 2, pp 149 Ae 158, 2024 been conducted to validate CHIRPS data.
The findings indicate that the accuracy of CHIRPS rainfall data in West Kalimantan is classified as very good .
However, the correlation between CHIRPS data and AWS data in South Lampung is in the weak category .
Additionally, the CHIRPS data correlation is relatively low due to topographic factors, the distance of rain stations to nearby mountains or oceans, and local wind circulation .
This research project aims to analyse historical changes in rainfall in the Ciliwung catchment area, with a particular focus on the capital city of Jakarta.
To achieve this, the research will validate CHIRPS satellite data against observational data based on previous research, namely changes in rainfall in the Ciliwung catchment area per year from 1993 to 2042, namely 1.
The objective is to gain an in-depth understanding of changes in rainfall in the Ciliwung watershed (Jakart.
and to assess the accuracy of satellite data in comparison to observational RESEARCH METHODOLOGY 1 Case Study: Ciliwung Watershed in Jakarta The study area is Ciliwung Watershed in Jakarta.
Indonesia.
The location of the Ciliwung Watershed in Jakarta is presented in Figure 1.
Jakarta is a densely populated area and has experienced significant land cover development.
The Ciliwung River is a river that flows in the DKI Jakarta area.
Bogor Regency.
Bogor City.
Depok City.
Bekasi, and surrounding areas.
Ciliwung is recorded as having a main stream length of 120 kilometers, while its catchment area .
iver flo.
is 387 km2.
The Ciliwung River Watershed has very strategic value because it crosses two provinces, namely West Java and DKI Jakarta, but due to the rapid development activities in these two provinces, it has caused significant land use changes and management.
The research was conducted in the Ciliwung River Watershed, especially the DKI Jakarta area, with an area of 13,995 ha.
This research covers the administrative areas of North Jakarta.
East Jakarta.
West Jakarta.
Central Jakarta, and South Jakarta in the Ciliwung watershed area.
Doi: https://doi.
org/10.
25105/livas.
Rainfall Correction Factor of Chirps Satellite Data Against Observation Data of Ciliwing Watershed (Case Study of Kemayoran Meteorological Statio.
Kurniyaningrum.
Faluty.
Mulya.
Andayani.
Hidayat.
Sejati.
Satar p-ISSN 2580-7552.
e-ISSN 2548-7515.
Volume 9.
Number 2, pp 149 Ae 158, 2024 Fig 1.
Ciliwung Watershed in Jakarta 2 Rainfall datasets The observed rainfall data presented here has been obtained from a rainfall observation station, namely the Kemayoran Meteorological Station.
The coordinates of the Kemayoran station are 106A 50' 24" East Longitude and 6A 9' 20" South Latitude.
This data includes observational rainfall data from the Indonesian Meteorological.
Climatological, and Geophysical Bureau (BMKG) in addition to baseline rainfall data from the Climate Hazards Group InfraRed Precipitation with Station (CHIRPS) project .
ttps://w.
edu/data/chirp.
spanning the past three decades, from 1993 to 2022.
Table 1.
Datasets Data Rainfall baseline (CHIRPS) Climate Projections (MICROC.
Rainfall observation Source https://w.
edu/data/chirps https://chelsa-climate.
org/future/ BMKG and BBWS Ciliwung Cisadane 3 Climate Hazard Group Infrared Precipitation with Stations (CHIRPS) CHIRPS (Climate Hazard Group Infrared Precipitation with Stations Dat.
is rainfall data that comes from combining observational data with satellite data, which has a spatial resolution of 05A .
er pixe.
or around 5kM x 5kM, to estimate sustainable changes in rainfall in a as well as for analysis of rainfall trends.
CHIRPS data availability began in 1981 and continues today with daily, monthly, and decade rainfall categories.
There are various factors that influence the accuracy of satellite rainfall data, such as environmental conditions, local Doi: https://doi.
org/10.
25105/livas.
Rainfall Correction Factor of Chirps Satellite Data Against Observation Data of Ciliwing Watershed (Case Study of Kemayoran Meteorological Statio.
Kurniyaningrum.
Faluty.
Mulya.
Andayani.
Hidayat.
Sejati.
Satar p-ISSN 2580-7552.
e-ISSN 2548-7515.
Volume 9.
Number 2, pp 149 Ae 158, 2024 climate, season, and topography.
CHIRPS data requires a further analysis process to validate the data against observed rainfall data.
4 Bias Correction The statistical method employed in bias correction as a re-analysis of observational rainfall data is multiple linear regression analysis with a stepwise method.
The initial step in correcting this method is to calculate the estimated rainfall according to the coordinates of the observation rainfall station, as illustrated by the following equation:
= yca yca1ycU1.
yca2ycU2.
yca3ycU3.
yca4ycU4.
= Rainfall observation day -i, grid -k yca = constant day -i, grid -k ycaycn = regression coefficient day -i ycU1.
= Rainfall day -i, grid -k ycU2.
= Grid longitude -k ycU3.
= Grid latitude -k The next step is to calculate the error value between the observed rainfall and the estimated ya.
= ycU.
Oe ycUA.
Then interpolate the error value according to the size and number of grids.
Next, calculate the estimated rainfall and error for all grids using equation .
And as the final step, calculate the corrected rainfall for the entire grid being analyzed by adding up the estimated rainfall with the error value.
Note that the corrected rain value, which is negative, is changed to zero.
The static performance test on CHIRPS Re-Analysis rainfall uses two static parameters, namely:
A Coefficient of Determination (R.
R2 shows the level of linear relationship between observation data and model data.
The R 2 value ranges from 0 to 1.
If R2 is 1, the result shows perfect agreement between the model data and the observed data.
Doi: https://doi.
org/10.
25105/livas.
Rainfall Correction Factor of Chirps Satellite Data Against Observation Data of Ciliwing Watershed (Case Study of Kemayoran Meteorological Statio.
Kurniyaningrum.
Faluty.
Mulya.
Andayani.
Hidayat.
Sejati.
Satar p-ISSN 2580-7552.
e-ISSN 2548-7515.
Volume 9.
Number 2, pp 149 Ae 158, 2024 ycI2 = [ ycuycayc OeycUycn iycuycayc ).
cUycn ycycnyco OeycUycn iycycnyco ) OcycA ycn=1[.
cUycn .
ycuycayc OeycUycn ycycnyco OeycUycn ocycA iycuycayc ) OcycA iycycnyco ) ycn=1[.
cUycn ycn=1[.
cUycn Where ycUIycnycuycayc is the average of observation data and ycUIycnycycnyco is the average of simulation data.
A NSE (Nash-Sutcliffe Efficienc.
NSE (Nash-Sutcliffe Efficienc.
represents how well the simulated value compares to the observed value.
The NSE value ranges from O to 1, and the closer to 1, the better the model performance is said to be.
NSE can be calculated using the equation (Gupta et al.
Ocycu .
cUycn ycuycayc OeycUycn ycycnyco ) ycAycIya = .
Oe ( ycn=1 ycu Ocycn=1.
cUycn ycuycayc OeycUIycn ycuycayc ) .
Table 2.
Parameter Assessment
Parameter Very Good
Good
Satisfactory Unsatisfactory NSE
75ONSEO1
65ONSEO0.
5ONSEO0.
NSEO0.
75OR O1
65OR2O0.
5OR2O0.
R2O0.
RESULTS AND DISCUSSION
Constraints in the availability of rainfall data in a region often become obstacles in research, mainly due to the uneven distribution of rain posts and a lack of complete observation data.
an effort to overcome this problem, this research utilizes global data sources, such as CHIRPS (Climate Hazard Group InfraRed Precipitation with Station Dat.
Fig 2.
Maximum Rainfall Pattern and Fitting Curve between Observed Rainfall and CHIRPS The CHIRPS rainfall data used in this research first carried out a performance test by comparing data obtained at rain posts in the Ciliwung watershed area of DKI Jakarta, namely the BMKG Meteorological Station.
Figure 2 shows the CHIRPS monthly maximum rainfall pattern and maximum observed rainfall, as well as the fitting curves at rain station.
Visually, the CHIRPS data Doi: https://doi.
org/10.
25105/livas.
Rainfall Correction Factor of Chirps Satellite Data Against Observation Data of Ciliwing Watershed (Case Study of Kemayoran Meteorological Statio.
Kurniyaningrum.
Faluty.
Mulya.
Andayani.
Hidayat.
Sejati.
Satar p-ISSN 2580-7552.
e-ISSN 2548-7515.
Volume 9.
Number 2, pp 149 Ae 158, 2024 rainfall pattern has rainfall values that are unable to follow observed rainfall.
Visually.
CHIRPS rainfall values are unable to follow observed rainfall.
This is supported by the static test results in Table 3, showing that the performance of CHIRPS data is very low, where R2 and NSE are in the not good enough category, or it can be said that CHIRPS data underestimates observational data.
So further analysis of CHIRPS rainfall values is needed by conducting trial and error.
Table 3.
CHIRPS Rainfall Static Test Results Kemayoran RA
NSE
Based on the R2 and NSE statistical tests, it shows that the performance of CHIRPS data compared to observation data is very low, where the R2 value is below 0.
5 and the NSE value is This shows that CHIRPS data underestimates observation data at selected rain So further analysis of CHIRPS rainfall values is needed by conducting trial and error.
After trial and error correction of bias 5 .
times, the CHIRPS data was tested for performance at the observation station.
The results show that the 5th trial has a good value.
The recap results in Table 4 show that the performance test has improved data quality compared to before analysis, namely that the R2 value shows a value above 0.
5, and likewise, with the NSE test, the corrected CHIRPS data for stations is close to 1, which indicates a more accurate model.
Table 4.
CHIRPS Rainfall Static Test Results (Tria.
Kemayoran RA
NSE
Figure 3 shows visually that the CHIRPS data after analysis shows good ability to follow patterns and approach the observed data values at both rain stations.
So it has great potential to be used to predict changes in rainfall in the future.
Fig 3.
Maximum Rainfall Pattern and Fitting Curve between Observed Rainfall and CHIRPS.
Doi: https://doi.
org/10.
25105/livas.
Rainfall Correction Factor of Chirps Satellite Data Against Observation Data of Ciliwing Watershed (Case Study of Kemayoran Meteorological Statio.
Kurniyaningrum.
Faluty.
Mulya.
Andayani.
Hidayat.
Sejati.
Satar p-ISSN 2580-7552.
e-ISSN 2548-7515.
Volume 9.
Number 2, pp 149 Ae 158, 2024 Then the CHIRPS data was compared by making a daily maximum rainfall curve from 1993 to 2022, both before and after correction.
This aims to assess how well MIROC5 accurately estimates actual rainfall and provides an important basis for further use of this data in future climate change-related research.
Analysis of baseline rainfall characteristics in the Ciliwung watershed, especially the Jakarta area, provides a significant picture of rainfall variability at monitoring stations.
Fig 4.
Box plot of maximum rainfall for the period 1993Ae2022.
Figure 4 displays a box plot of baseline rainfall at Kemayoran Station in the Ciliwung watershed, providing additional insight into understanding the climate characteristics of the Ciliwung The maximum average annual rainfall for 30 years .
3Ae2.
87 mm/year.
the minimum average is 227.
47 mm/year.
while the average annual rainfall at Kemayoran Station 32 mm/year.
CONCLUSION
CHIRPS data estimates for daily rainfall for the period 1993Ae2022 at the BMKG meteorological station tend to have very good CHIRPS data accuracy values .
verage percent bias = 9.
NSE
= 0.
This shows that CHIRPS data is able to estimate events of rain in the Ciliwung watershed area, especially in the Jakarta area.
CHIRPS estimates are higher in light rainfall (< 30 m.
and lower in heavy to very heavy rainfall.
Utilization of CHIRPS data can be done by first increasing the accuracy of CHIRPS data by applying a correction factor to two groups of daily Doi: https://doi.
org/10.
25105/livas.
Rainfall Correction Factor of Chirps Satellite Data Against Observation Data of Ciliwing Watershed (Case Study of Kemayoran Meteorological Statio.
Kurniyaningrum.
Faluty.
Mulya.
Andayani.
Hidayat.
Sejati.
Satar p-ISSN 2580-7552.
e-ISSN 2548-7515.
Volume 9.
Number 2, pp 149 Ae 158, 2024 rainfall categories, namely the light category and the medium to very heavy category in the research area.
This is supported by the land cover conditions in the research area.
CHIRPS data can be categorized as very good if used as an alternative to limited observational rainfall data, which can then be used in analyzing water availability in the Ciliwung watershed (Jakart.
ACKNOWLEDGEMENT
The authors convey appreciation to thank to Balai Besar Wilayah Sungai Ciliwung-Cisadane for providing all the necessary data for this study and encouragement to conduct such studies for the benefit of science and society.
REFERENCES