Undertakes researches on tropical cyclones and other disaster causing phenomena and on weather modification.Health Risk Map Related to Tropical Cyclone Occurrence in Metro Manila
The impact on health condition during tropical cyclone occurrences in Metro Manila was studied in an attempt to detrmine the storm's influence on disease activity, and in delineating vulnerable areas.
The study concentrated on seven (7) landfalling tropical cyclones and hospital admission of eight diseases (gastroenteritis, dengue fever, hepatitis, influenza, measles, typhoid fever, bronchopneumonia, and tetanus) for the period 1984 to 1988. Two-tailed ANOVA at 5% level of significance was used o isolate diseases affected by the storm's passage. Health risk maps were developed to delineate vulnerable areas.
The result showed that hospital admission of dengue fever cases increased 1-and 2-month after the cyclone passages while gastroenteritis, hepatitis, typhoid fever, bronchopneumonia and tetanus admission decreases after the typhoon. On the other hand, the meteorological disaster showed no significant effect on the activity of influenza and measles.
Human vulnerable areas in the Metropolis were identified in the health risk maps developed and can be very useful in health mitigation. Further study is recommended with focus on incorporation of health data from other hospitals.
Results of the analysis of observed extreme wind speeds in stations throughout the country are illustrated and dicussed. For clarity of the discussion extreme wind was categorized, depending on the increasing magnitude of the wind speed, into moderately extreme (60 to 100 kph), extreme (101 to 184 kph) and severely extreme (185 or more).
The study shows that severely extreme wind speeds of 185 kph and above are experienced in several areas of the country during various months of the year. In the region of Luzon, the areas are namely, Itbayat, Basco, Vigan, Aparri, Tuguegarao, Dagupan, Munoz, Casiguran, Port Area, Tayabas, Ninoy Aquino International Airport (NAIA), Ambulong, Infanta, Alabat, Daet and Legaspi. The affected areas in the Visayas region are Virac, Romblon and Masbate, Catarman, Tacloban, Guiuan and Mactan; while in the Mindanao region, extreme wind speeds are observed in the northernmost portion, i.e. in Surigao del Norte.
Being an archipelagic country with 7,100 islands located at the so called "typhoon alley of the world", the Philippines is exposed to very strong winds especially during the passage of tropical cyclones. hence, it is deemed necessary for the government to undertake measures as that would help its seventy five million Pilipinos to undertake preparedness measures as well as structural and non-structural mitigation strategies to minimized the effects of extreme wind hazards.
On the average, the annual average of tropical cyclones forming over the North Western Pacific Ocean side of the Philippine Area of Responsibility (PAR) is 20. Hence, forecasters are generally familiar with their characteristics. However, the same can not be said of tropical cyclones originating from the South China Sea. In September 1998, Tropical Storm Gading formed in the South China Sea and seriously affected Western Pangasinan. As a result, a fact-finding board recommended that extensive study should be made for tropical cyclones originating from South China Sea. Initial assessments showed that these tropical cyclones have very unusual motions and quite rare (52 tropical cyclones in 50 years). Out of 52 tropical cyclones that formed, only 13 made landfall in the Philippines.
The present study aims to do a climatological analysis of the database so as to explain the cyclogenesis and behavior of the storm in the area. The findings could provide indicators or precursors that may be useful in forecasting the formation and movement of storms in the South China Sea.
As an initial effort, the objective techniques at the NDRB were used to come up with methods that may help forecasters in predicting storms in the said area. Forecast verification of tropical cyclone movement using the objective forecasting technique was performed. Results show that Persistence with Climatology [(P+C)/2] method gave that least error or bias.
A quantitative rainfall forecasting technique was develop through the use of output simulation of a two-dimensional Primitive equation Model (PEM) of tropical cyclone rainfall (Villareal, 1999). Maximum rainfall and radius of maximum rainfall from the simulation were used to developed a rain profile. The rain profile was patterned after a wind profile equation (Sulapat et al, 1996) which was originally developed by Holland (1980). The rain profile was tested and applied to eight cases of troipcal cyclones that landfall in the Philippine Area of Responsibility (PAR) from 1990 to 1997. The results of application showed that the model profile can reasonably predict the location of maximum rainfall. However, the model needs further improvements in terms of accuracy of rainfall prediction.
To improve the above forecasting technique, simple statistical procedure was applied between the observed and forecast rainfall to develop a simple regression equation on each tropical cyclone case. This procedure is simple called Model Output Statiitics (MOS). Results showed that the regression equation for Rosing exhibits the highest R2 value of 0.692. To assess the foreacasting capability of MOS, the equation was applied to cases of tropical cyclone rainfall that landafall in 1998 (Katring and Loleng). Results of MOS application for Klaring showed that MOS also reasonably predict the location and amount of maximum rainfall on August 7 and 8, 1998. MOS application for Loleng showed that the location and amount of maximum 24-hr rainfall were not accurately predicted.
Flood haard mapping and vulnerability analyses conducted in eight coastal towns of Bataan are presented. The main criteria used in the selection of the study areas, are their geographical location along the Manila Bay and existence of historical flooding. Thus, the focus of study are the coastal towns namely, Hermosa, Abucay, Balanga, Limay, Samal, Orani, and Pilar.
The study aims to develop a database on flooding and vulnerability of these coastal towns in Bataan, and to evaluate the potential of flood occurrence. It also hopes to correlate the physical social and economic conditions that make the area vulnerable to flooding and lastly, to prepare flood hazard and vulnerability maps of these coastal towns in Bataan. Secondary data used include the historical data on flooding, barangay maps, topography or contour maps, soil maps and hydrographic maps. The primary data consist of actual survey on flooding and on-site observation of recent flood events with still photographs taken in the study areas.
A simple hazard mapping method was used to delineate the flood prone areas. Analysis consists of defining the extent of flooding directly from the recorded inundation areas and data obtained from actual on-site surveys of recent flooding. In the absence of hydrologic and hydraulic data in Bataan, the evaluation was done by relating flooding to the primary physical characteristics of the coastal municipalities of Bataan using topographical and geographical information.
In general terms, the result of the study indicate that the geographical extent and severity of flooding are affected by the geomorphology, physiography, water run-off, topography and inadequate drainage system. Specially, the coastal towns, namely, Hermosa, Abucay, Balanga, Samal, Orani, Orion and Pilar are most prone to flooding. The geographical delineation of flood vulnerable areas show that these are highly vulnerable to flooding as a consequence of their loction in the lowland areas along the lower reaches of Manila de Bay. Furthermore, these places are densely populated with marginalized, vulnerable people. The agricultural lands, fish ponds and critical facilities or lifelines are locted in the flood prone areas., hence making them highly vulnerable to disaster especially during the southwest monsoon season. Only the town of Limay has a verylow susceptibility to flooding, due to higher elevtion, being closer to the foot of the Mariveles Mountain.
Socio-economic influence on human response to tropical cyclone warning was studied to assess the impact of typhoon warning as a tool to disaster mitigation. Human response survey was conducted in five provinces affected by T. Gading, T. Iliang and T. Loleng, which occurred in 1998.
Out of 687 respondents, majority will not be taking any action regardless of gender, educational attainment, age, civil status, family density, career level, family income, type of houses or even warning acceptance. Regression result show that except for family income, which has no significance to human response, other socio-economic vaiables have minimal influences and still have inverse relationship with response.
The result also reveals that there are 468 respondents who claim full comprehension of the warning message. Verification result, however, show that there are only 28 respondents who understand the warning.
The result suggests the need for simplification of the warning, improvement of public education methodology on natural hazards, and promulgation of disaster mitigation compliance policy.
The evaluation of the predictive skills of various objective methods used in tropical cyclone forecasting by the Philippine Atmospheric Geophysical Astronomical Services Administration (PAGASA) is discussed. The study specifically analyzed and compared the forecast performance of the objective forecasting techniques using the data on tropical cyclone movement from 1994 to 1998.
Prior to this study, a preliminary evaluation was done using the forecasts results from tropical cyclones for the period covering 1994 to 1996, utilizing various methods, namely, Persistence, Climatology, Amadore 2, Persistence cum Climatology [(P+C)/2], Analog, Barotropic Model and the Model Output Statistics (MOS). The tropical cyclone forecast data were stratified by month of occurrence and by geographical location of cyclone passage. The results of the forecast evaluation (based on the least error compared to the tropical cyclone best track) were used as the basis for the choice of objective techniques to be used in the formulation of the official PAGASA forecast.
However, in the present study, forecasts from the Barotropic Model and the MOS were not included due to non-availability of sufficient data to make a meaningful evaluation. Among the forecasting techniques mentioned above, Persistence cum Climatology [(P+C)/2] was more frequently used than the other methods.
The evaluation of the forecast skill of the chosen predictive techniques was determined using the data of tropical cyclones from 1997 to 1998 in the validation, in order to avoid bias. Results show that the chosen forecasting methods gave smaller average errors than the other techniques. Furthermore, there were greater percentages of "correct" forecasts or "hits" (59%) than "errors" or "misses" as compared to the official forecast.
Vorticity, Omega and Wind Analysis at the 850 hPa level of the FLM12 (24-Hr Forecast) were studied during the occurrence of Typhoon Gloria and Herb over the Northwestern Pacific from 20 July to 01 August 1996 at 00 UTC. The derived Maximum Positive Vorticity (MPV), Maximum Negative Omega (MNO) and cyclonic wind circulation from the analyzed charts that were considered as the centers of the circulation of the disturbances in the study, were compared with the surface real-time position. While the analyses were able to produce centers of circulation of the two disturbances, the MPV's and MNO's actual location differs with the real-time position. However, the result of the Wind Analysis at this level is encouraging especially in tropical cyclone tracking.
A two dimensional theoretical model capable of describing the radial distribution of tropical cyclone rainfall is adapted. The model uses primitive equation written in cylindrical coordinate. The grid system is non-uniform with a smaller grids located near the center and at the surface to give emphasis to the eye of the tropical cyclone. Representing the drag coefficient with a constant value modifies the model.
The results of the modified model simulation show many realistic features typical of a matured tropical cyclone. It realistically simulated the inflow at lower levels, the upward motion near the center and the outflow at upper levels. On the other hand, the simulated rainfall rate intensity is found to be linear with respect to tropical cyclone intensity. Likewise, the present model determines the location of maximum rainfall rate and maximum wind. It shows that the location of maximum rainfall is closer to the center than the location of maximum wind regardless of tropical cyclone intensity.
Hazard maps showing regions delineated by "isobronts" i.e. lines of equal thunderstorm (TSTM) activity in the Philippines are presented. The data used in the study consist of TSTM observations during a ten-year period covering 1984 to 1993. Data stratification was done according to time of TSTM occurrence and according to season. Such stratification was deemed necessary in order to take into account in the analysis, the effect of three (3) important air streams affecting the country namely, the Northeast Monsoon, the Southwest Monsoon and the North Pacific Trades.
The objectives of the study among others are: 1) To develop a database on TSTM occurrence in the Philippines, 2) To delineate areas using "isobronts" i.e. lines of equal thunderstorm activity and 3) To prepare a TSTM hazard map showing areas prone to TSTM in the Philippines.
Results of the study indicate that during the northeast monsoon season, the peak of thunderstorm activity is at 8:00 p.m. local time (1200 UTC). The time at which TSTMs attain peak occurrence during the southwest monsoon season is at 5:00 p.m.(0900 UTC). During the transition period, TSTM occur most frequently at 0900 UTC (5:00 P.M.) and at 1200 UTC (8:00 P.M.). The greatest number of TSTMs occur during the southwest monsoon season, followed by the northeast monsoon and transition months.
During the northeast monsoon season TSTMs most frequently occur at the western areas in the country. Areas along the western coast and at the central portions are more subjected to TSTM during the southwest monsoon season and transition period. Out of the 60,532 thunderstorms observed in the country over the 10-year period from 1984 to 1993, about 7,558 TSTM (12.48%) occurred during the Northeast Monsoon season, 45,462 (75.4%) during the Southwest Monsoon season, and 7,392 (12.21%) during the Transition period.
This paper deals with the modification of the analog method in 1998. The main feature of the study is the use of a personal computer in software of the method including the construction of the probability ellipses which may guide forecasters in their official forecasts.
An attempt to improve the analog method is performed through the introduction of ensemble forecasting. The noise introduced in the initial data seems not to improve the method.
Results show a skill of the analog that is similar to the analog of 1980's and is at times better than other objective forecasting techniques.
Two methods of rainfall estimation were developed and computed for three (3) tropical cyclones of 1995. The first method was the Satellite Technique that followed the method of Scofied and Oliver (1980) using satellite imagery. The second method was the Modified Composite Technique where the use of composite rainfall analysis of J.F. Bucoy, et. al. (1987) was employed. The techniques were developed through computer programming to forecast the 6,12,18 and 24 hr rainfall. The application of the methods required forecast of the 24-hourly position of the tropical cyclone.
The results of the present applications showed that the average error of the forecast positions were-42 km for latitude and 44 km for longitude. The negative (positive) value of forecast error means that the forecast position is behind (ahead) the observed position. On the basis of the resulting estimate the corresponding rainfall of Satellite Technique Experiment Number 1 had the smallest mean average error of 18.91%, while Satellite Technique Experiment Number 3 had the largest mean average error of 254.30%. However the Modified Composite Technique had mean average error of 80.20%. The verification showed that Satellite Technique Experiment Number 1 performed better over Satellite Technique Number 2. Number 3 and the Modified Composite Technique. The results of the model were encouraging but it needs further improvements, calibration and modification.
The conducted study was to develop area rainfall forecasting scheme for Metro Manila using surface data from three weather stations in the area.
The combination of three analytical strategies was considered in the formation of forecasting method. These are regression and climatology (F1), climatology and persistence (F2), regression and persistence (F3), regression, climatology and persistence (F4), and the adjusted regression and climatology (F5).
The result showed that for cyclone that crossed Metro Manila F1 and F5 may be used. However, for cyclone that does not landfall but is just passing Northeast of Metro Manila, the model needed the aid of other forecasting tools.
Improvement of the method may be studied using upper air data or the adaptation of meso-scale analysis of the surface data.
The development and application of the Direct Model Output Statistics (DMOS) method to improve the forecast of tropical cyclone tracks is discussed. The forecast positions of tropical cyclones which entered the Philippine Area of Responsibility (PAR) for the period 1991 to 1993 consisting of 234 forecast cases were utilized as the development data set for the study.
The initial latitude and longitude positions of tropical cyclones, and the forecasts of the latest modified operational Barotropic Model (2x2 grid system) were taken as predictors, while the predictands included the 12-hourly forecasts. The multi-linear regression equations from both the 500 hPa level and deep layer mean (DLM) data of objectively analyzed maps, were utilized as predictive equations and were the bases of the new Direct Model Output Statistics (DMOS) forecasts. Forecast verification was undertaken using 171 tropical cyclone forecast cases which occurred from 1994 to 1996.
Results of the study show the remarkable usefulness of the Direct Model Output Statistics (DMOS) in forecasting tropical cyclone tracks. Percentages of improvement of the forecasts of the Barotropic Model using the DMOS method are substantially high. The over-all latitude forecast improvement, using DMOS-500 was 13.2% for the 24-hr, 30.4% for the 48-hr and 39.6% for the 72-hr forecast periods. While the over-all longitude forecast improved by as much as 26.9%, 42.1% and 61.4% for the 24th, 48th, and 72-hour forecast periods respectively.
A health-weather relationship for Metro Manila was studied. The analysis considered cholera, dengue, malaria, measles, meningitis, and typhoid fever including the equivalent weather variables temperature, relative humidity, and rainfall. Graphical and statistical correlations were conducted to explore the underlying epidemiological facts beneficial to health disaster planning and mitigation. Lagged variables are also included.
The result discovered some predictive capability of the three weather variables for cholera, dengue, malaria, and measles epidemic one to three months before the outbreak. Further investigation is encouraged for meningitis and typhoid fever.
Hazard maps of thunderstorm frequencies from 1984 -1993 of 15 synoptic stations over the Visayas are presented. Three (3) important air streams affecting the Philippines namely: Northeast Monsoon, the Southwest Monsoon and the North Pacific Trades were taken into account in the analyses. The study revealed that of the 18,709 thunderstorms observed over the 10-year period, 10% occurred during the NE Monsoon Season, 18% during the Transition Period and 72% during the SW Monsoon Season. Statistically, this shows that the western parts of Visayas are more vulnerable to thunderstorm hazard. Moreover, findings of the study revealed that thunderstorms attained peak occurrences at 0900 UTC (5:00 pm) during both the NE and SW Monsoon Seasons as compared to the transition Period (North Pacific Trades) where the peak occurrence of thunderstorm is at 0600 UTC (2:00 pm).
Hazard maps showing thunderstorm and lightning frequencies from 26 synoptic stations over Luzon for period covering 1984 - 1993 are presented. The relationship of thunderstorm activity to the three (3) important air streams affecting the Philippines namely: Northeast Monsoon, the Southwest Monsoon and the North Pacific Trades is discussed.
The study indicates that occurrences of thunderstorms and lightnings are more frequent during Southwest Monsoon season that during the Northeast Monsoon and the North Pacific Trades. The TSTM and lightning hazard maps show that the western parts of Luzon are more vulnerable to these natural hazards. Moreover, the statistical data indicate that the maximum frequency of thunderstorms and lightnings occurrence is between 06Z - 12Z and at its minimum during the rest of the day.
The characteristics of southwest monsoon surge was investigated during the period from 1981 to 1987 considering the pressure difference of selected stations, such as: Iloilo; Cuyo Island; Puerto Prinsesa; Coron Island; PAGASA Island; Sangley Point; and San Jose with reference to Port Area station, Manila. Based on 116 cases, the results showed that pressure difference was generally lowest in August and highest in April. April revealed no southwest monsoon surge during this month and most frequent during August. Favorable time for southwesterly onset was at 0000 UTC (8:00 AM) and the least favorable time was at 0009 UTC (5:00 PM). The typical DP values of Iloilo station (637) of surge onset range from -1.0 to -2.0 hPa. The minimum values occurred some 3-hours prior to the surge arrival at an average of -1.7 hPa.
The data of three tropical cyclone cases, namely; Typhoon Katring in 1994; Tropical Storm Mameng and Typhoon Rosing in 1995 were analyzed for winds, warnings and damages. The severe weather bulletins, maximum sustained winds (MWS), passage reports, barograph charts, summary of damages and some pictures were also collated.
The results of the analysis showed that cases that are over warned, correctly warned and under warned in terms of Modified Public Storm Warning System (MPSWS) were 33%, 45% and 225 respectively. Also, the collected pictures of damages demonstrated good patterns of recorded winds brought about by these three tropical cyclone cases.
This paper describes techniques applied to rainfall prediction for metro Manila using surface data taken from NAIA Weather station. The analytical scheme consisted of stepwise regression with the inclusion of climatology. Three (3) cases were processed (Case 1 - unstratified, Case 2 - Southwest Monsoon, Case 3 - Northeast Monsoon).
Verification of result showed that the combined effect of regression analysis and climatology (Case 1, equation 2) gave the highest percentage accuracy for heavy rain prediction (30,8%) and 48.5% for moderate rain prediction. However, both are still below 50%. Light rain forecast have high percentage accuracy for all forecasting scheme developed ranging from 70.2% to 98.6%.
Suggestion is made to apply the developed forecasting scheme to data of Science Garden and Port Area.
The geographical delineation of flood vulnerable areas in Taguig is discussed. The study area covers the 18 barangays in Taguig situated in the eastern part of Metro Manila and the northwestern shore of Laguna de Bay.
Results show that those areas that are highly vulnerable to flooding are located in the lowland areas and along the coast of Laguna de Bay and other major rivers in Taguig. Some of these areas are densely populated, others are agricultural lands and while others have the concentration of critical facilities or lifelines in Taguig.
A simple approach to flood hazard mapping is discussed. The study area (43.3 square Kilometers) is the municipality of Taguig, Metro Manila which is located in the northwestern shore of Laguna de Bay, about 15 Km. East of the city of Manila.
The techniques consist of defining the flood hazard directly from recorded inundation areas and actual on-site- surveys of recent floodings. The evaluation was done by relating the flood to the primary physical characteristics of Taguig utilizing topographical and geographical information. This alternatives approach was pursued due to the non-availability of hydrologic and hydraulic data for Taguig. Verification and validation of the prepared maps was undertaken through the interview of residents in the area.
Results show that the extent and severity of flooding in Taguig is affected by the geomorphology, physiography, water run-off, topography and improper drainage.
The surface wind profiles, following basically that of Holland's model (1980) were constructed for ten landfalling tropical cyclones cases from 1993 to 1995. Analysis of ten cases revealed that the surface maximum wind speed of 93.5 kilometers-per-hour was 36.8 kilometers from the tropical cyclone center, on the average. The computed parameter ▀which determined the shape of the surface wind speed profile was of the order of 2.15.
A proposed nomogram forecasting technique in rainfall prediction applicable for Pinatubo area was constructed.
Sixteen year historical surface date of Iba Weather Station during tropical cyclone incident in the area bounded by 5-250 N latitude, 117-1270 E longitude was used as independent data set to obtain a rainfall nomogram for the Pinatubo area Performance of the method was tested and verified using five tropical cyclone cases of 1993.
The experiment showed that in some instance, the method is efficient but at other times it fails. The error was of the order 0.6-26.0 millimeters.