Utilization of Himawari-8 Satellite Data for Estimated Rainfall with Various Methods (Case Study of Heavy Rain Jayapura 16-18 March 2019)

The estimation of rainfall is very important in increasing the accuracy of the weather forecast. An accurate weather forecast will reduce the impactof disaster risks. Jayapura was chosen as the stufy are because in recent times there was heavy rain
of 8
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
    Utilization of Himawari-8 Satellite Data for Estimated Rainfall with Various Methods (Case Study of Heavy Rain Jayapura 16-18 March 2019) A R Sagala 1 , H Salawane 2 , and A K Silitonga 3   Meteorological Climatological and Geophysical Agency (BMKG) 1,2 Angkasa I No.2 Kemayoran Street, 10720 DKI Jakarta, Indonesia, School of Meteorology Climatology Geophysics (STMKG) 3 Perhubungan I No.5 street, 15221 Tangerang Selatan, Banten, Indonesia Abstract. The estimation of rainfall is very important in increasing the accuracy of the weather forecast. An accurate weather forecast will reduce the impactof disaster risks. Jayapura was chosen as the stufy are because in recent times there was heavy rain for three consecutive days which caused various losses. Estimation of rainfall  by applying various estimators have been widely carried out but specifically discussing the phenomenon of heavy rainfall is still small. This research was carried out by utilizing the Himawari-8 channel 13 (Infrared1) satellite data as input to estimate rainfall and compared to Authomatic Weather Station Digitation Dok II Jayapura rainfall data. The estimator methods used are Autoestimator, Insat Multispectral Rainfall Algorithm (IMSRA), Non-Linear Relation, Non-Linear Inversion. Cloud top temperature data for three heavy rain events (16-18 March 2019) are converted into rainfall data every hour during the convection process. Phase of cloud develompment (growing, mature, extinct) based on time series cloud top temperature which indicates a decreace in temperature at growing  phase, low maximum in mature phase, and increase in extinct phase. The result of this study show that the best estimation method for heavy rain in Jayapura base on ranking point (higest correlation and lowest RMSE) is IMSRA. 1.   Introduction The auto estimator method is applied to estimate daily rainfall in Jayapura in January-February 2010. The results of the study indicate that in general the auto estimator method is relatively poorly used for estimating rainfall in Jayapura but is good for estimating the number of rainy days. The auto estimator method has been widely used to estimate the amount of rainfall. However, the results of rainfall estimates detected by weather satellites and funds on synoptic observations have a difference or value difference. This is what needs to be assessed for the accuracy of the estimated rainfall using this method. Conventional rain rainfall observation encountered many obstacles, especially when bad weather occurred. This is a problem for the availability of timely, continuous and accurate rainfall data. Therefore, the estimation of rainfall using weather satellite imagery needs to be developed 1 . Rainfall prediction in Indonesia is still very difficult to accurately predict because it has very large spatial and temporal variations with the number of surface observation rainfall observations which is still very limited 2 . The study of heavy rain cases in Jayapura in 2014 showed a significant decrease in the value of cloud top temperatures around -61C and a significant increase in cloud top height. These values and  patterns can be used as an indication of the growth of heavy rain-producing clouds in the Jayapura region 3 .   Precipitation patterns over sea surface should be more easily estmated than over land since the  patterns are probably prompted more by purely atmospheric factors and leey by complex topographic variation, yielding instantaneous distributions of precipitation activeity that are more random than those over land, but average distribution that are less complex 4 .    When viewed from the trigger, the shape of the cloud caused by local conditions or the usual shear in the form of one cell cumuliform clouds. Whereas regions, where there are ITCZ or vortex, low, tropical depression, in satellite images, will be seen the formation of cumuliform clouds mixed with stratiform clouds. In this case, the resulting rain will usually have a long and continuous period, which will eventually result in an accumulation of large rainfall. For flood-prone areas, this can be a problem  because when the accumulation of rainfall exceeds the extreme weather threshold of the region, flooding will occur  5 . Remote sensing is a science and measurement method to get information on an object or  phenomenon using a recording device from a distance without making physical contact with objects or measured/observed phenomena 6 . A wheater satellite, which has wider coverage and faster in generating data than in situ observation, is used to monitor weather in Indonesia to provide atmospheric dynamic information that is related to early warning of extreme weather. It can generate cloud top brightness temperature data to determine the presence and type of clouds, and it also can be used to generate the amount of rainfall 7 . 2.   Data and Method The area of study in this study is Jayapura 2 Meteorological Station. The coordinates of this region are 2.53S 140.72E. The characteristic of the Jayapura region is the hills with many green areas. The Dok 2 Jayapura Meteorological Station is right on the coast of Jayapura waters so that the phenomenon of land wind and sea wind can be one of the causes of extreme weather in Jayapura. ( Figure 1 . Map of Jayapura source: thematic map of indonesia ) The data used in this research are: 1.   Himawari 8 sattelite data on 16-18 March 2019 Infrared (IR1) channel with 10 minutes resolution with data extension .Z. Data for per 10 minutes for time series cloud top temperatures and per hourly for estimated rainfall.   2.   Hourly rainfall data from Authomatic Wheater Station Digitation Dok 2 on 16-18 March 2019.   Cloud top temperature is extracted from the satellite IR channel Himawari-8. The data is then entered into various equations that convert cloud top temperatures into rainfall in this case used 5 equations that have been studied in existing research. Precipitation estimates from various equations are compared to the AWS Digitization Dock 2 rainfall. From these comparisons, the correlation and RMSE values will  be obtained. The incidence of heavy rain for three days in a row was compared separately as well as the    correlation and RMSE. The ranking table of 3 events will be used to choose the best estimation method with the highest correlation and lowest RMSE. The frequency of the appearance of CB clouds in Serui, Papua in March 182 events in 2014 and most occurred at 15:00 UTC as many as 26 events. The emergence of CB clouds in Yapen is caused by land wind convergence from the island of Papua and the island of Yapen. This condition lasts from 11:00 UTC to 16:00 UTC.   8 The Himawari 8 and 9 satellites are advanced programs from the MTSAT Satellite, to improve the accuracy of observation and weather prediction developed into 16 bands consisting of 3 visible bands, 3 near-infrared bands (NIR) and 10 thermal bands (IR) with spatial resolution 0 , 5km x 0.5km and 1km x 1km per pixel on visible band data, 2km x 2km per pixel on IR bands and 1km x 1km and 2km x 2km per pixel on the NIR band so the data volume becomes large. Besides, the intensity of observations increases to every 10 minutes, so the data transmission needs to be faster and the volume  becomes large. For this reason, JMA will develop a fast data receiver system and for user needs, AHI data will be distributed with a resolution directly via the internet. The increasing number of bands and the increasing temporal and spatial resolution of the Himawari 8 and 9 Satellite data, is very necessary and is expected to provide an opportunity to improve the accuracy of weather prediction in tropical regions such as Indonesia which has rapid weather change dynamics 8 . Himawari 8 and 9 have special meteorological missions, namely: maintaining continuity and increasing weather observations via satellite for disaster prevention and weather forecasting; improve the ability of short-term forecast the next 6 hours or what is known as 'nowcasting', especially for the detection and prediction of bad weather; improve the accuracy of numerical weather predictions; improve climate and environmental monitoring 9 .   2.1.   Autoestimator Comparison of GEOS cloud-top temperature in the IR images and visible images against collocated radar images have demonstrated that convective thunderstorms are characterized by very low clud-top temperature (195-210 K) and rapid spatial and temporal changes in the structure of the cloud-top surface. The auto-estimator initially computes rainfall rates based on a nonlinear, power-law regression relationship between cloud-top temperature (10.7 um brightness temperature) and radar-derived rainfall estimates.  = 1,1183.10  exp (3,6382.10 − . . ) (2.1) Autoestimator uses National Centers for Environmental Prediction (NCEP) Eta Model generated relative humidity (RH) and precipitable water (PW) to analyze the environmental moisture and scale the rainfall amounts accordingly. The auto-estimator has some skill at 1-h time resolution and spatial resolutions of 12 km, and it presents useful results on larger grid sizes (48-km and larger). Cold-top (colder than 280 K) MCSs composed the dataset for this initial validation 10 . Contrast to the reasonale performace of the technique for well-defined and short duration convective systems, poor results are common for stratiform cloud systems. This is especially important during winter season when most of the precipitation comes from cloud tops that are warmer than 230 K. For these case, danger of flas flooding result from persistent rainfall over long periods of time, not to short-lived intense storms associated with convective systems during the summer. These intense convective systems are the ones for which the auto-estimator was srcinally developed. We should also be carefull about the use of auto-estimator for 24-h daily rainfall accumulation. 2.2.   IMSRA The IMSRA technique is combination of the IR and MW measurements which benefits from the relative accuracy of the MW-based estimates and the relatively low sampling errors of the TIR-based estimates. This algoithms is developed for the small-scale rainfall estimation over the Indian region.  = 8,613098∗  ( (−7,7)5,7  )  (2.2)    The development of this algorithm includes two major steps: classification of rain bearing clouds using  proper clud classification scheme utilizing Kalpana-1 TR and WV Tbs; collaction of Kalpana-1 IR Tbs with TRMM-PR suface rainfall rate and establishment of a regression relation between them. Kalpana-1 based rainfall estimate is successfullynused for rainfall monitoring during the monsoon and severe weather conditions 11 . 2.3.    NonLinear Relation Information on rainfall is very important because it is needed by various aspects of life, especially in  planning agriculture, transportation, plantations to early warnings of old disasters, floods/landslides and droughts. Remote sensing satellites for weather and environmental monitoring can provide weather information every hour in a wide range. The lower the brightness temperature of the cloud, the higher the rainfall, except for cirrus clouds which are not rain-producing but have low temperatures 12 .  = 2.10 5 . ,5   (2.3)  2.4.    NonLinear Inversion The relationship between the physical properties of the object under study with observational data can  be in the form of a mathematical model. This model is used to extract the physical parameters of the object under study from observation data. This process is called inverse modelling 13 .  = 1,38046.10 −7 . ,   (2.4) The rainfall model obtained using nonlinear inversion based on the cloud top temperature, seen from the comparison graph between the observed rainfall and the estimated rainfall has a relatively similar  pattern, but the magnitude of the increase or decrease in rainfall is not the same. This is caused by other factors such as cloud top temperature factors that affect rainfall such as latitude, altitude, a distance of water sources, differences in soil temperature and land area. 3.   Result and Discussion (a) (b)    (Figure 2 . cloud top temperature (a) March 16, 2019, (b) March 17, 2019, (c) March 18, 2019) Based on the time series of cloud top temperatures determined the time series of cloud development at the study point, namely at the Jayapura II Meteorological Station. March 16, 2019, the time series which is used as a phase of cloud development is 09.00 UTC-20.00 UTC (12 data). March 17, 2019, the time series as a cloud phase is 12.00 UTC-21.00 UTC (10 data). March 18, the time series used is 12.00 UTC-20.00 UTC (9 data). The choice of time series of this cloud development phase will indirectly filter out other hours where the chance of rain is small because the development of convection clouds is a common cause of rain. Using AWS data can also actually determine the phase of cloud development. High cloud top temperature correlates with high rainfall and generally occurs in the mature phase. Rainfall tops occur at very low cloud top temperatures (mature phase) while in the growing and extinct phases of rainfall are relatively small changed in cloud top temperatures in these two phases are opposite. Based on cloud top temperatures it is determined that the growing phase of clouds on 16, 17, March 18, 2019, is 09.00-09.30 UTC, 12.00-13.30 UTC, 12.00-14.00 UTC. The mature phase of the cloud on 16, 17, 18 March 2019 is 9:30 to 17:00 UTC, 13.30-18.00 UTC, 14.00-19.00 UTC. The extinct phase of the cloud on the 16th, 17th, March 18, 2019, is 17.00-20.00 UTC, 18.00-21.00 UTC, 19.00-21.00 UTC.   (Figure 3. comparison of rainfall on March 16, 2019) The rainfall at 14:00 UTC is the highest rainfall at 40.8 mm. At this hour, the Autoestimator overestimate method is 43.1 mm against the actual rainfall while the rainfall estimation using the Non-Linear Relation method is closest to the actual rainfall despite underestimation with a rainfall value of 38.3 mm. The Autoestimator method experiences the greatest overestimate at 12, 13, and 16 UTC, which is the phase 0.050.0100.091011121314151617181920    R   a   i   n    f   a    l    l    (   m   m    /    h   r    ) UTC Comparative Rainfall Estimation Jayapura 16-03-2019 AEIMSRANL RelationNL InversionAWS (c)
Similar documents
View more...
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks

We need your sign to support Project to invent "SMART AND CONTROLLABLE REFLECTIVE BALLOONS" to cover the Sun and Save Our Earth.

More details...

Sign Now!

We are very appreciated for your Prompt Action!