SlideShare a Scribd company logo
1 of 14
Michael Stavely
DIGITAL IMAGE PROCESSING FINAL PROJECT
The Size of Tropical Cyclones Crossing Madagascar into the
Mozambique Channel Measured through Precipitation Correlated
with the El Nino Southern Oscillation Teleconnection
1
Introduction:
Tropical Cyclones crossing over Madagascar into the Mozambique Channel typically
diminish due to the orographic lifting and subsequent intense rainfall on the island of
Madagascar. Some of these storms maintain Tropical Cyclone strength, while others diminish to
less intense classifications of storms. There appears to be a correlation between the warming of
the sea surface in the local waters associated in the timing of the peak La Nina events occurring
during the El Nino Southern Oscillation (ENSO) and stronger and more frequent Tropical
Cyclones. In order to better understand the impact of the El Nino Southern Oscillation on
Tropical Cyclones of the Southwest Indian Ocean and the Mozambique Channel, precipitation
data was used to measure the size of the cyclones once they pass over Madagascar in order to
determine if they maintain tropical cyclone strength, and if there is a relationship between ENSO
values and strength. This data can aid in prediction of strength of storms and their impact on the
African mainland once they cross Madagascar. This will also reinforce the impact of the ENSO
teleconnection on the energy in the meteorological system on the area. There is not many studies
of the area in regards to Tropical Cyclones that cross Madagascar and the extra knowledge of the
phenomena as well as any association with ENSO will be helpful in predicting the longevity and
severity of Tropical Cyclones which fit this profile.
ResearchQuestion:
Is there an association with the strength of tropical cyclones as determined by size of the
cloud formation defined by a 5 millimeters of rainfall per hour boundary with the teleconnection
El Nino Southern Oscillation as rated by severity indices?
2
Study Area:
This study was conducted in the
Mozambique Channel between the African
nations of Mozambique on the mainland
and the island of Madagascar. In Figure 1
this area is highlighted in yellow to
distinguish the boundaries of the study.
Also highlighted in Figure 1 are the tracks
of tropical cyclones in the seasons studied
of 2000, 2003, and 2004 while they were
involved within the study area. As
described in the image, the 2000 season
tracks are represented in red and
correspond with high La Nina conditions,
the 2003 season storm track is represented in blue and corresponds with high El Nino conditions;
and the 2004 season storm tracks are represented in green and corresponds with “normal”
conditions.
While relatively few Tropical Cyclones cross over Madagascar and enter the
Mozambique Channel, there appears to be patterns of groups of strong storms as well as groups
of storms not strong enough to be considered Tropical Cyclones. The mountainous regions of
Madagascar tend to disrupt the storms eliminating the weak bands and only allowing the stronger
bands and core to remain above the 5 millimeter per hour precipitation parameter for Tropical
Cyclones.
Figure 1 shows the study area with the six qualifying storms as they
crossed the Mozambique Channel which is defined by the yellow
polygon.
3
Data:
For this project Tropical Rainfall Measuring Mission (TRMM) precipitation data was
used in order to determine the strength of the Tropical Cyclones in the study area. TRMM comes
in multiple temporal ranges including monthly, and daily averages as well as a 3-hour hourly
average with up to 1-hour products available through computation. To preserve the accuracy of
the individual time sections because Tropical Cyclones are relatively fast moving meteorological
formations, 3-hour product was downloaded for each of the dates specified from the track
interaction with the study area which provides data in millimeters per hour. This product can be
used to calculate the cyclone strength pixels as any pixel that registers greater than 5 millimeters
per hour can be classified as Tropical Cyclone strength. In order to increase the processing, only
the images from 0z were downloaded and processed. Limitations with the data include the
unreliability of TRMM in measuring precipitation over land. Also, TRMM data is only recorded
from 1997 and reliable from 1998, so previous years data is not available and accurate change
over time is difficult to fully describe. Another limitation of TRMM is the large amount of
preprocessing required to make the data usable. In order to run analysis, the TRMM data must be
downloaded, unzipped and loaded into ENVI as an HDF. Next, the pixel visualization must be
defined, the data must be rotated, and the geographic data must be defined before the data can be
viable. The data is available online through NASA.
Image 1 shows an example Tropical Rainfall Measuring Mission (TRMM) data set between the 35N and 35S latitudes. This
product is a 3 hour average which the pixel values are displayed in millimeters per hour. For this project this data was
downloaded at 0z for every date which the selected storms interacted with the study area.
4
In order to determine qualifying storms for the project, an El Nino Southern Oscillation
Severity index was required to select which years were appropriate to examine. Figure 2 was
used to determine ideal years to obtain data for as a strong El Nino (peak red value), strong La
Nina (peak blue value), and normal condition year was required for comparison. Due to the
limitation of TRMM data that it is only valid from 1998 to 2014, the seasons 2003 and 2010
were identified for strong El Nino conditions; the seasons 2002 and 2004 were identified for
“normal” conditions; and the seasons 2000, 2008, and 2011 were identified for strong La Nina
conditions. Of these seasons, two storms fulfilled the study area and tropical cyclone strength
requirements during El Nino conditions, two storms fulfilled the requirements for “normal”
conditions, and six storms fulfilled the requirements during La Nina conditions. In the El Nino
years, Joel, the only storm of the 2010 season began as a subtropical storm, so the season
examined for El Nino was 2003. For the “normal” conditions, 2002 was devoid of storms
fulfilling the requirements, and thus the 2004 season was chosen. Also of note, the 2004 storm
Figure 2 shows El Nino Southern Oscillation values rated by sea surface temperature anomaly with areas classified in red as El
Nino years and blue as La Nina years.
5
Gafilo was the strongest storm of record to form in the Indian Ocean. As for the strong La Nina
conditions, the 2000 season was chosen to represent because of its intensity of conditions and its
frequency of storms.
Figures 3 and 4 were also use to predict the frequency and strength of the storms in
relation to El Nino and La Nina conditions. Note that in Figure 3 which depicts the sea surface
temperature anomaly during El Nino conditions that the warm water temperatures is limited to
the upper part of the study area. In contrast, in Figure 4 the warm waters are shown surrounding
much of the island of Madagascar which suggests that the waters the storms enter before
crossing Madagascar and entering into the study area provides the storms with energy to pass
over the orographic landscape and retain Tropical Cyclone strength.
Figure 4 shows global sea surface temperatures as opposed
to the "normal" conditions during an El Nino year with an
emphasis on the East Indian and West Pacific Oceans.
Figure 4 shows global sea surface temperatures as opposed
to the "normal" conditions during a La Nina year with an
emphasis on the East Indian and West Pacific Oceans.
6
Table 1 shows the six storms examined with their dates of relevance to the study area in both Y-M-D, as well as Julian date. Also
included is the duration in days of each storms activity in the study area. This was calculated through IBTrACS data.
Name Year Month Day Hour (UTM) Julian Duration (days)
Start Astride 1999 12 31 12 365 4
Stop 2000 1 3 6 3
Start Leon-Eline 2000 2 17 6 48 6
Stop 2000 2 22 12 53
Start Hudah 2000 4 2 12 92 7
Stop 2000 4 8 12 98
Start Japhet 2003 2 24 12 55 8
Stop 2003 3 3 0 62
Start Cela 2003 12 9 12 343 11
Stop 2003 12 19 18 353
Start Gafilo 2004 3 6 18 65 8
Stop 2004 3 13 6 72
Table 1 describes the chosen storms with the interaction with the study area. It includes
the start time and the end time of the storms track within the study area. Also included is the date
in both Year-Month-Day, as well as the Julian calendar. The time listed is the 3-hour product
when the storm track intersected with the study area. The duration is calculated through the
difference in the Julian days as well as the addition of the start day. This was used to select the
TRMM data dates to download at 0z for these dates. The addition of the extra day regardless of
the hour of interaction should make the hour insignificant in order to maintain uniform data. In
order to obtain the dates, IBTrACS data was loaded into ArcMap and the highlighted storm
tracks were selected out of the total. Next, the study area was defined as a polygon and a
selection by location was performed to extract the beginning and end of the tracks involvement
with the study area. These dates were then exported and compiled into Table 1. The final results
of the track selection lines can be examined in Figure 1.
7
Methods:
The first step is to use the ENSO Index to decide which seasons are appropriate for
displaying the peak values of El Nino, La Nina, and “normal”. More information about this
process was discussed in the Data section and refer any questions there. The selected storms
which attain Tropical Cyclone Strength during their lifetime and cross over the island nation of
Madagascar during the selected seasons of 2000, 2002, 2003, 2004, 2008, 2010, and 2011 are
listed below. Of these the seasons and storms that were chosen include 2000, 2003, and 2004:
Season Storms ENSO Category
2000 Astride, Leon-Eline, Hudah La Nina
2002 N/A “Normal”
2003 Japhet El Nino
2004 Cela, Gafilo “Normal”
2008 Ivan, Jokwe La Nina
2010 Joel El Nino
2011 Giovanna La Nina
Table 2 shows the seasons and storms that correspond with the peak ENSO condition diplayed in the table. The table is also
color co-ordinated based on the ENSO category. Of these seasons, 2000, 2003, and 2004 were chosen to represent their
categories.
The next step involves loading the TRMM data into ENVI for raster manipulation. One
of the biggest challenges for this project is the sheer extent of the data temporally and the
numerous factors that can affect the storm strength. Of the selected storms, TRMM data from
every date as determined by Table 1 at 0z was downloaded and organized by season and storm.
Next the raw data was preprocessed. The preprocessing of the data is one of the most time
consuming factors and the TRMM data can be difficult to load properly. However, all analysis is
useless if the data is not properly displayed as the project compares the rainfall to the study area,
but the locations do not line up. From NASA there is a detailed list to properly load the data:
8
TRMM L-3 HDF:| (GES DISC; NASA)
1. From the command line, type in, ENVI,
2. Click File|Open External File|Generic Format|HDF. Select a file and click OK
3. Select both precipitation and relativeError (e.g., 3B43.20100201.7.HDF) and click OK.
Select BSQ in HDF Data Set Storage Order
4. Select Basic Tools|Rotate/Flip Data. Select Data Set #1 (precipitation) and click OK. Type
in -90 degrees and select File.
5. Right click the new file and select Edit Header. Click Edit Attributes|Map Info. Fill in the
table with Image X = 1, Image Y = 1; -180.0 and 50.0 for Map Coordinate of the Tie Point.
Put 0.25 degree for Pixel Size of both X and Y. Click OK.
6. Pick the parameter you just added the map info to and click Load Band.
Once the TRMM data has been properly loaded, a spatial subset should be performed
around the study area in order to focus on the location. Afterwards, for each band perform a
Color Slice in ENVI from the table of contents and remove all the current color slices. Add a
new slice and define the minimum at 5 and leave the maximum. This creates a selection of only
the pixels which are valued at 5 millimeter per hour precipitation or greater which can then be
exported to a shapefile and loaded into ArcMap. This can be measured in respect to the study
area to determine the amount of space in the study area which is valued at Tropical Cyclone
strength. This can be quantified and analysis can be run to compare the results to the ENSO
severity index to determine correlation between the two.
9
Results:
The results of the methods are compiled in Figures 5, 6, and 7 and represent the 2004
“normal” conditions, the 2003 El Nino conditions, and the 2000 La Nina conditions respectively.
Figure 5 which represents the “normal” conditions shows the two storms Cela and Gafilo along
with their tracks and the shapefiles which represent their pixel of precipitation which exceed 5
millimeters per hour. As you can see from the figure, both storms registered a fairly large amount
of pixels on the maps and also followed similar paths coming from the northeast of the image
and crossing the island, and then proceeding to rotate back across Madagascar. The data and the
track that are green correspond to Gafilo, and the data and track that are red correspond to Cela.
Gafilo was the strongest recorded storm in the Southwest Indian Ocean.
Figure 6 shows strong El Nino conditions and is paired with the track and precipitation
data of Japhet in green as well as the track of Joel in blue with crosses. The precipitation data for
Joel was not run due to its origin as subtropical, but since the El Nino season only included one
Figure 5 shows the Tropical Cyclone Tracks
for the 2000 season which is associated with
strong La Nina conditions.
Figure 7 shows the Tropical Cyclone Tracks
for the 2003 season which is associated with
strong El Nino conditions. Joel track is
included only to show track. Data associated
with the Japhet is also included.
Figure 6 shows the Tropical Cyclone Tracks
for the 2004 season which is associated with
"normal" conditions. Precipitation data for
Cela and Gafilo is represented as well with
red and green respectively.
10
Tropical Cyclone, the Joel track was added for comparison. Based on these two tracks, El Nino
years also tend to show a similar pattern moving from around the southern part of the channel
across into the mainland. Also, neither storm directly crossed over Madagascar. The origin of
Japhet is also a point of interest as it is the only Tropical Cyclone in the study which originated
within the channel itself. Despite all of its precipitation data being present in the image and its
lack of crossing the mountainous terrain of Madagascar, very few pixels are highlighted as
Tropical Cyclone strength in the channel. This could be due to the TRMM 3-hourly data
recorded which was then averaged to obtain millimeters per hour, but overall, Japhet was one of
the weakest storms recorded through the study.
Figure 7 shows the 2000 season
which was used to represent the La Nina
conditions. The tracks of the three storms
Astride, Leon-Eline, and Hudah are
displayed on the figure. All three seem to
have a uniform pattern of crossing
Madagascar on the northern half of the
island, the dipping south upon channel
reentry and then moving onto the
mainland. Precipitation data of Leone-
Eline, the strongest storm in the channel of
this study, is displayed in Figure 8. This is
the strongest storm due to much of Gafilo
TRMM values being recorded over land.
Figure 8 shows the precipitation data as the Tropical Cyclone moves
across the Mozambique Channel. Feb 19 and 22 produced the most
catagorized rainfall.
11
Conclusions:
In conclusion, based on the years sampled and the method of quantifying the storms,
there does seem to be a connection between occurrence and strength of Tropical Cyclones which
cross into the Mozambique Channel and the El Nino Southern Oscillation, specifically, years of
low ENSO (La Nina) tends to have very active Tropical Cyclone seasons with the strongest
storm in the channel as well as the most storms in the sampling years at six while the other two
conditions created two storms each. Also in support of this is the strong El Nino conditions lack
of creating a storm which crossed the island of Madagascar with one originating in the channel
and the other forming as a subtropical storm. Also, during “normal” conditions, while there were
two storms formed, they both occurred in the same year while the other values recorded no
storms. And finally, Gafilo was such a strong storm that the precipitation display could easily be
linked to other factors external of the study area.
Discussion:
Some of the limitations of the data include the inability to isolate the El Nino Southern
Oscillation as the sole factor in the storm creation. Other teleconnections such as the Madden-
Julian Oscillation (MJO) may have had an impact on the storms. Further research should focus
on the sea surface temperatures of the individual storms in the region of formation and follow
through to their entry into the channel. The limitation of TRMM not being reliable over land
would require land precipitation gauges to be incorporated into the data, or the selection of
another form of precipitation data. The inability to create an extensive profile due to the temporal
range of TRMM is also a limiting factor. Another factor is the rating of when the storms in
12
question were recorded as Tropical Cyclones. All storms selected were rated at Tropical
Cyclones during their lifetimes, but may not have been even before beginning to cross
Madagascar. Due to the large amounts of images to be processed, only selecting the 0z time of
the 3-hour TRMM images may have removed higher value images from the test pool, and so a
more regular set of images may be required to fully understand the data. Another factor was the
size of the storms, since most were too small to see the eyewall and core cloud formation for
measurement and tracking purposes. Another question raised by this project was if there seems
to be uniform path the Tropical Cyclones seem to follow as they cross the channel as determined
by their ENSO category as the current tracks seems to suggest. And finally, by transforming the
color slice of the raster directly into a shapefile before recording the values, it was not possible to
easily ascertain the number of pixels valued at greater than 5 in the study area. This limited the
amount of quantitative analysis that was able to be performed on the results. This project was a
very large scope in order to examine a general climatologic trend, but for more accurate and
definitive result, more intimate studies of each storm in question need to be performed in order to
understand the impact of external factors on the final data.
13
References:
El Niño and La Niña Years and Intensities. (n.d.). Retrieved December 4, 2015, from
http://ggweather.com/enso/oni.htm
ENSO Myth Number 4 - The Variations in the East Pacific and the East Indian-West Pacific Sea
Surface Temperatures Counteract One Another. (2013, March 27). Retrieved December 4, 2015,
from http://wattsupwiththat.com/2013/03/27/enso-myth-number-4-the-variations-in-the-east-
pacific-and-the-east-indian-west-pacific-sea-surface-temperatures-counteract-one-another/
How can I import TRMM data in ENVI? (n.d.). Retrieved December 11, 2015, from
http://disc.sci.gsfc.nasa.gov/recipes/?q=faq/How-can-I-import-TRMM-data-ENVI
Index of TRMM Data. (n.d.). Retrieved from
ftp://disc2.nascom.nasa.gov/data/s4pa/TRMM_L3/TRMM_3B42/
Matyas, C. (n.d.). International Journal of Climatology Volume 35, Issue 3, Article first
published online: 7 APR 2014. Retrieved December 4, 2015, from
http://onlinelibrary.wiley.com/doi/10.1002/joc.3985/pdf
TRMM Tropical Cyclones - SurfaceRain Tutorial. (n.d.). Retrieved December 11, 2015, from
http://www.nrlmry.navy.mil/sat_training/tropical_cyclones/trmm/surfacerain/
Tropical Rainfall Measurement Mission. (n.d.). Retrieved December 11, 2015, from
http://trmm.gsfc.nasa.gov/
Vitart, F., Anderson, D., & Stockdale, T. (2003, June 19). 3932 V OLUME 16 JOURNAL OF
CLIMATE q 2003 American Meteorological Society Seasonal Forecasting of Tropical Cyclone
Landfall over Mozambique. Retrieved December 4, 2015, from
http://journals.ametsoc.org/doi/pdf/10.1175/1520-0442(2003)0162.0.CO;2

More Related Content

What's hot

The atmosphere 1 ESO
The atmosphere 1 ESOThe atmosphere 1 ESO
The atmosphere 1 ESOmihayedo
 
John Gallant_Topographic parameters for Australia
John Gallant_Topographic parameters for AustraliaJohn Gallant_Topographic parameters for Australia
John Gallant_Topographic parameters for AustraliaTERN Australia
 
Earthquake findings
Earthquake findingsEarthquake findings
Earthquake findingsshelbyteresa
 
An expert system model for identifying and mapping tropical wetlands and peat...
An expert system model for identifying and mapping tropical wetlands and peat...An expert system model for identifying and mapping tropical wetlands and peat...
An expert system model for identifying and mapping tropical wetlands and peat...CIFOR-ICRAF
 
Sea level rise impact modelling on small islands: case study gili raja island...
Sea level rise impact modelling on small islands: case study gili raja island...Sea level rise impact modelling on small islands: case study gili raja island...
Sea level rise impact modelling on small islands: case study gili raja island...Luhur Moekti Prayogo
 

What's hot (6)

The atmosphere 1 ESO
The atmosphere 1 ESOThe atmosphere 1 ESO
The atmosphere 1 ESO
 
John Gallant_Topographic parameters for Australia
John Gallant_Topographic parameters for AustraliaJohn Gallant_Topographic parameters for Australia
John Gallant_Topographic parameters for Australia
 
Earthquake findings
Earthquake findingsEarthquake findings
Earthquake findings
 
An expert system model for identifying and mapping tropical wetlands and peat...
An expert system model for identifying and mapping tropical wetlands and peat...An expert system model for identifying and mapping tropical wetlands and peat...
An expert system model for identifying and mapping tropical wetlands and peat...
 
Sea level rise impact modelling on small islands: case study gili raja island...
Sea level rise impact modelling on small islands: case study gili raja island...Sea level rise impact modelling on small islands: case study gili raja island...
Sea level rise impact modelling on small islands: case study gili raja island...
 
Wilks_et_al_2016
Wilks_et_al_2016Wilks_et_al_2016
Wilks_et_al_2016
 

Viewers also liked

Rural service delivery. framin draft1
Rural service delivery. framin draft1Rural service delivery. framin draft1
Rural service delivery. framin draft1Elias Elwai Williams
 
Presentacion andragogía ppt 2016
Presentacion andragogía ppt 2016Presentacion andragogía ppt 2016
Presentacion andragogía ppt 2016Digna De Puy M.
 
Presentacion sobre manejo de materiales
Presentacion sobre manejo de materialesPresentacion sobre manejo de materiales
Presentacion sobre manejo de materialessantiago mariño
 
Evolución de las partes internas del computador
Evolución de las partes internas del computadorEvolución de las partes internas del computador
Evolución de las partes internas del computadorNeyda Viana
 
Programa de historia_de_la_arquitectura_i
Programa de historia_de_la_arquitectura_iPrograma de historia_de_la_arquitectura_i
Programa de historia_de_la_arquitectura_iNeyda Viana
 
Marlon mendoza c.i. 25140613
Marlon mendoza c.i. 25140613Marlon mendoza c.i. 25140613
Marlon mendoza c.i. 25140613marlon mendoza
 
Hjerneskaderehabilitering 2
Hjerneskaderehabilitering 2Hjerneskaderehabilitering 2
Hjerneskaderehabilitering 2Jonna Midtgaard
 
David_Curruculum_Vitae
David_Curruculum_VitaeDavid_Curruculum_Vitae
David_Curruculum_VitaeDavid Prakash
 
I vicini di casa. Convivere e condividere è possibile
I vicini di casa. Convivere e condividere è possibileI vicini di casa. Convivere e condividere è possibile
I vicini di casa. Convivere e condividere è possibileAnna Carrara
 
Pheonix-Design-Studio
Pheonix-Design-StudioPheonix-Design-Studio
Pheonix-Design-StudioSudhir Rao
 
SINET_EdNexus_BeyondTeacherEvaluations_071614
SINET_EdNexus_BeyondTeacherEvaluations_071614SINET_EdNexus_BeyondTeacherEvaluations_071614
SINET_EdNexus_BeyondTeacherEvaluations_071614Kathleen T. Hayes, Ed.D.
 
Short break Berlin SEPT14
Short break Berlin SEPT14Short break Berlin SEPT14
Short break Berlin SEPT14Lollie Barr
 
Proposta c.c. n. 2 2016
Proposta c.c. n. 2 2016Proposta c.c. n. 2 2016
Proposta c.c. n. 2 2016NOMENSAUNICA
 
Weekend Villa near Ahmedabad
Weekend Villa near AhmedabadWeekend Villa near Ahmedabad
Weekend Villa near AhmedabadSuryam Developers
 
Drilling Supervisor and Engineer with more than 12 years of experience
Drilling Supervisor and Engineer with more than 12 years of experienceDrilling Supervisor and Engineer with more than 12 years of experience
Drilling Supervisor and Engineer with more than 12 years of experienceNILANJAN PODDAR
 

Viewers also liked (20)

Rural service delivery. framin draft1
Rural service delivery. framin draft1Rural service delivery. framin draft1
Rural service delivery. framin draft1
 
Presentacion andragogía ppt 2016
Presentacion andragogía ppt 2016Presentacion andragogía ppt 2016
Presentacion andragogía ppt 2016
 
Presentacion sobre manejo de materiales
Presentacion sobre manejo de materialesPresentacion sobre manejo de materiales
Presentacion sobre manejo de materiales
 
Evolución de las partes internas del computador
Evolución de las partes internas del computadorEvolución de las partes internas del computador
Evolución de las partes internas del computador
 
Programa de historia_de_la_arquitectura_i
Programa de historia_de_la_arquitectura_iPrograma de historia_de_la_arquitectura_i
Programa de historia_de_la_arquitectura_i
 
SME Concept. Draft1
SME Concept. Draft1SME Concept. Draft1
SME Concept. Draft1
 
Marlon mendoza c.i. 25140613
Marlon mendoza c.i. 25140613Marlon mendoza c.i. 25140613
Marlon mendoza c.i. 25140613
 
Hjerneskaderehabilitering 2
Hjerneskaderehabilitering 2Hjerneskaderehabilitering 2
Hjerneskaderehabilitering 2
 
teacher-evaluations-and-local-flexibility
teacher-evaluations-and-local-flexibilityteacher-evaluations-and-local-flexibility
teacher-evaluations-and-local-flexibility
 
Actividad tecnologica
Actividad tecnologicaActividad tecnologica
Actividad tecnologica
 
David_Curruculum_Vitae
David_Curruculum_VitaeDavid_Curruculum_Vitae
David_Curruculum_Vitae
 
I vicini di casa. Convivere e condividere è possibile
I vicini di casa. Convivere e condividere è possibileI vicini di casa. Convivere e condividere è possibile
I vicini di casa. Convivere e condividere è possibile
 
Pheonix-Design-Studio
Pheonix-Design-StudioPheonix-Design-Studio
Pheonix-Design-Studio
 
Dinusha CV with project
Dinusha CV with projectDinusha CV with project
Dinusha CV with project
 
SINET_EdNexus_BeyondTeacherEvaluations_071614
SINET_EdNexus_BeyondTeacherEvaluations_071614SINET_EdNexus_BeyondTeacherEvaluations_071614
SINET_EdNexus_BeyondTeacherEvaluations_071614
 
Short break Berlin SEPT14
Short break Berlin SEPT14Short break Berlin SEPT14
Short break Berlin SEPT14
 
Proposta c.c. n. 2 2016
Proposta c.c. n. 2 2016Proposta c.c. n. 2 2016
Proposta c.c. n. 2 2016
 
Weekend Villa near Ahmedabad
Weekend Villa near AhmedabadWeekend Villa near Ahmedabad
Weekend Villa near Ahmedabad
 
BoettcherCaseStudy-Final
BoettcherCaseStudy-FinalBoettcherCaseStudy-Final
BoettcherCaseStudy-Final
 
Drilling Supervisor and Engineer with more than 12 years of experience
Drilling Supervisor and Engineer with more than 12 years of experienceDrilling Supervisor and Engineer with more than 12 years of experience
Drilling Supervisor and Engineer with more than 12 years of experience
 

Similar to Final Paper

ArdhiArbain-LightningGPM20140114
ArdhiArbain-LightningGPM20140114ArdhiArbain-LightningGPM20140114
ArdhiArbain-LightningGPM20140114Ardhi Adhary Arbain
 
Modelling the Typhoon Haiyan Storm Surge on Eastern Philippines Using Weather...
Modelling the Typhoon Haiyan Storm Surge on Eastern Philippines Using Weather...Modelling the Typhoon Haiyan Storm Surge on Eastern Philippines Using Weather...
Modelling the Typhoon Haiyan Storm Surge on Eastern Philippines Using Weather...Nathan Hipolito
 
Paper id 71201909
Paper id 71201909Paper id 71201909
Paper id 71201909IJRAT
 
Tr 102 adamson_1981_southern_african_storm_rainfall_nodata
Tr 102 adamson_1981_southern_african_storm_rainfall_nodataTr 102 adamson_1981_southern_african_storm_rainfall_nodata
Tr 102 adamson_1981_southern_african_storm_rainfall_nodataPieterSteenkamp10
 
Identification of three_dominant_rainfall_regions_
Identification of three_dominant_rainfall_regions_Identification of three_dominant_rainfall_regions_
Identification of three_dominant_rainfall_regions_Lasriama Siahaan
 
L021203080089
L021203080089L021203080089
L021203080089theijes
 
CEPSI 2014 Full paper JKT Alstom WIND TURBINE OPERATION IN TYPHOON CONDITIONS
CEPSI 2014 Full paper JKT Alstom WIND TURBINE OPERATION IN TYPHOON CONDITIONSCEPSI 2014 Full paper JKT Alstom WIND TURBINE OPERATION IN TYPHOON CONDITIONS
CEPSI 2014 Full paper JKT Alstom WIND TURBINE OPERATION IN TYPHOON CONDITIONSJosef Tadich
 
diurnal temperature range trend over North Carolina and the associated mechan...
diurnal temperature range trend over North Carolina and the associated mechan...diurnal temperature range trend over North Carolina and the associated mechan...
diurnal temperature range trend over North Carolina and the associated mechan...Sayem Zaman, Ph.D, PE.
 
EVALUATION OF VERTICAL REFRACTIVITY PROFILE OVER MICROWAVE LINK IN MOWE, NIGERIA
EVALUATION OF VERTICAL REFRACTIVITY PROFILE OVER MICROWAVE LINK IN MOWE, NIGERIAEVALUATION OF VERTICAL REFRACTIVITY PROFILE OVER MICROWAVE LINK IN MOWE, NIGERIA
EVALUATION OF VERTICAL REFRACTIVITY PROFILE OVER MICROWAVE LINK IN MOWE, NIGERIAjmicro
 
EVALUATION OF VERTICAL REFRACTIVITY PROFILE OVER MICROWAVE LINK IN MOWE, NIGERIA
EVALUATION OF VERTICAL REFRACTIVITY PROFILE OVER MICROWAVE LINK IN MOWE, NIGERIAEVALUATION OF VERTICAL REFRACTIVITY PROFILE OVER MICROWAVE LINK IN MOWE, NIGERIA
EVALUATION OF VERTICAL REFRACTIVITY PROFILE OVER MICROWAVE LINK IN MOWE, NIGERIAjmicro
 
Hydrometeorological Hazard Maps in the Philippines.ppt
Hydrometeorological Hazard Maps in the Philippines.pptHydrometeorological Hazard Maps in the Philippines.ppt
Hydrometeorological Hazard Maps in the Philippines.pptMARJONMVILLONES
 
Hydrometeorological Hazard Maps in the Philippines.ppt
Hydrometeorological Hazard Maps in the Philippines.pptHydrometeorological Hazard Maps in the Philippines.ppt
Hydrometeorological Hazard Maps in the Philippines.pptMARJONMVILLONES
 
Study of Average Hourly Variations of Radio Refractivity Variations across So...
Study of Average Hourly Variations of Radio Refractivity Variations across So...Study of Average Hourly Variations of Radio Refractivity Variations across So...
Study of Average Hourly Variations of Radio Refractivity Variations across So...iosrjce
 

Similar to Final Paper (20)

Nwp final paper
Nwp final paperNwp final paper
Nwp final paper
 
ArdhiArbain-LightningGPM20140114
ArdhiArbain-LightningGPM20140114ArdhiArbain-LightningGPM20140114
ArdhiArbain-LightningGPM20140114
 
#5
#5#5
#5
 
SevereTCGeorge
SevereTCGeorgeSevereTCGeorge
SevereTCGeorge
 
Modelling the Typhoon Haiyan Storm Surge on Eastern Philippines Using Weather...
Modelling the Typhoon Haiyan Storm Surge on Eastern Philippines Using Weather...Modelling the Typhoon Haiyan Storm Surge on Eastern Philippines Using Weather...
Modelling the Typhoon Haiyan Storm Surge on Eastern Philippines Using Weather...
 
Paper id 71201909
Paper id 71201909Paper id 71201909
Paper id 71201909
 
Tr 102 adamson_1981_southern_african_storm_rainfall_nodata
Tr 102 adamson_1981_southern_african_storm_rainfall_nodataTr 102 adamson_1981_southern_african_storm_rainfall_nodata
Tr 102 adamson_1981_southern_african_storm_rainfall_nodata
 
Identification of three_dominant_rainfall_regions_
Identification of three_dominant_rainfall_regions_Identification of three_dominant_rainfall_regions_
Identification of three_dominant_rainfall_regions_
 
F43012934
F43012934F43012934
F43012934
 
L021203080089
L021203080089L021203080089
L021203080089
 
Iau solar effects 2005
Iau solar effects 2005Iau solar effects 2005
Iau solar effects 2005
 
CEPSI 2014 Full paper JKT Alstom WIND TURBINE OPERATION IN TYPHOON CONDITIONS
CEPSI 2014 Full paper JKT Alstom WIND TURBINE OPERATION IN TYPHOON CONDITIONSCEPSI 2014 Full paper JKT Alstom WIND TURBINE OPERATION IN TYPHOON CONDITIONS
CEPSI 2014 Full paper JKT Alstom WIND TURBINE OPERATION IN TYPHOON CONDITIONS
 
Ll3519561960
Ll3519561960Ll3519561960
Ll3519561960
 
diurnal temperature range trend over North Carolina and the associated mechan...
diurnal temperature range trend over North Carolina and the associated mechan...diurnal temperature range trend over North Carolina and the associated mechan...
diurnal temperature range trend over North Carolina and the associated mechan...
 
2-04_BPPT_Arbain-Galihselowati
2-04_BPPT_Arbain-Galihselowati2-04_BPPT_Arbain-Galihselowati
2-04_BPPT_Arbain-Galihselowati
 
EVALUATION OF VERTICAL REFRACTIVITY PROFILE OVER MICROWAVE LINK IN MOWE, NIGERIA
EVALUATION OF VERTICAL REFRACTIVITY PROFILE OVER MICROWAVE LINK IN MOWE, NIGERIAEVALUATION OF VERTICAL REFRACTIVITY PROFILE OVER MICROWAVE LINK IN MOWE, NIGERIA
EVALUATION OF VERTICAL REFRACTIVITY PROFILE OVER MICROWAVE LINK IN MOWE, NIGERIA
 
EVALUATION OF VERTICAL REFRACTIVITY PROFILE OVER MICROWAVE LINK IN MOWE, NIGERIA
EVALUATION OF VERTICAL REFRACTIVITY PROFILE OVER MICROWAVE LINK IN MOWE, NIGERIAEVALUATION OF VERTICAL REFRACTIVITY PROFILE OVER MICROWAVE LINK IN MOWE, NIGERIA
EVALUATION OF VERTICAL REFRACTIVITY PROFILE OVER MICROWAVE LINK IN MOWE, NIGERIA
 
Hydrometeorological Hazard Maps in the Philippines.ppt
Hydrometeorological Hazard Maps in the Philippines.pptHydrometeorological Hazard Maps in the Philippines.ppt
Hydrometeorological Hazard Maps in the Philippines.ppt
 
Hydrometeorological Hazard Maps in the Philippines.ppt
Hydrometeorological Hazard Maps in the Philippines.pptHydrometeorological Hazard Maps in the Philippines.ppt
Hydrometeorological Hazard Maps in the Philippines.ppt
 
Study of Average Hourly Variations of Radio Refractivity Variations across So...
Study of Average Hourly Variations of Radio Refractivity Variations across So...Study of Average Hourly Variations of Radio Refractivity Variations across So...
Study of Average Hourly Variations of Radio Refractivity Variations across So...
 

Final Paper

  • 1. Michael Stavely DIGITAL IMAGE PROCESSING FINAL PROJECT The Size of Tropical Cyclones Crossing Madagascar into the Mozambique Channel Measured through Precipitation Correlated with the El Nino Southern Oscillation Teleconnection
  • 2. 1 Introduction: Tropical Cyclones crossing over Madagascar into the Mozambique Channel typically diminish due to the orographic lifting and subsequent intense rainfall on the island of Madagascar. Some of these storms maintain Tropical Cyclone strength, while others diminish to less intense classifications of storms. There appears to be a correlation between the warming of the sea surface in the local waters associated in the timing of the peak La Nina events occurring during the El Nino Southern Oscillation (ENSO) and stronger and more frequent Tropical Cyclones. In order to better understand the impact of the El Nino Southern Oscillation on Tropical Cyclones of the Southwest Indian Ocean and the Mozambique Channel, precipitation data was used to measure the size of the cyclones once they pass over Madagascar in order to determine if they maintain tropical cyclone strength, and if there is a relationship between ENSO values and strength. This data can aid in prediction of strength of storms and their impact on the African mainland once they cross Madagascar. This will also reinforce the impact of the ENSO teleconnection on the energy in the meteorological system on the area. There is not many studies of the area in regards to Tropical Cyclones that cross Madagascar and the extra knowledge of the phenomena as well as any association with ENSO will be helpful in predicting the longevity and severity of Tropical Cyclones which fit this profile. ResearchQuestion: Is there an association with the strength of tropical cyclones as determined by size of the cloud formation defined by a 5 millimeters of rainfall per hour boundary with the teleconnection El Nino Southern Oscillation as rated by severity indices?
  • 3. 2 Study Area: This study was conducted in the Mozambique Channel between the African nations of Mozambique on the mainland and the island of Madagascar. In Figure 1 this area is highlighted in yellow to distinguish the boundaries of the study. Also highlighted in Figure 1 are the tracks of tropical cyclones in the seasons studied of 2000, 2003, and 2004 while they were involved within the study area. As described in the image, the 2000 season tracks are represented in red and correspond with high La Nina conditions, the 2003 season storm track is represented in blue and corresponds with high El Nino conditions; and the 2004 season storm tracks are represented in green and corresponds with “normal” conditions. While relatively few Tropical Cyclones cross over Madagascar and enter the Mozambique Channel, there appears to be patterns of groups of strong storms as well as groups of storms not strong enough to be considered Tropical Cyclones. The mountainous regions of Madagascar tend to disrupt the storms eliminating the weak bands and only allowing the stronger bands and core to remain above the 5 millimeter per hour precipitation parameter for Tropical Cyclones. Figure 1 shows the study area with the six qualifying storms as they crossed the Mozambique Channel which is defined by the yellow polygon.
  • 4. 3 Data: For this project Tropical Rainfall Measuring Mission (TRMM) precipitation data was used in order to determine the strength of the Tropical Cyclones in the study area. TRMM comes in multiple temporal ranges including monthly, and daily averages as well as a 3-hour hourly average with up to 1-hour products available through computation. To preserve the accuracy of the individual time sections because Tropical Cyclones are relatively fast moving meteorological formations, 3-hour product was downloaded for each of the dates specified from the track interaction with the study area which provides data in millimeters per hour. This product can be used to calculate the cyclone strength pixels as any pixel that registers greater than 5 millimeters per hour can be classified as Tropical Cyclone strength. In order to increase the processing, only the images from 0z were downloaded and processed. Limitations with the data include the unreliability of TRMM in measuring precipitation over land. Also, TRMM data is only recorded from 1997 and reliable from 1998, so previous years data is not available and accurate change over time is difficult to fully describe. Another limitation of TRMM is the large amount of preprocessing required to make the data usable. In order to run analysis, the TRMM data must be downloaded, unzipped and loaded into ENVI as an HDF. Next, the pixel visualization must be defined, the data must be rotated, and the geographic data must be defined before the data can be viable. The data is available online through NASA. Image 1 shows an example Tropical Rainfall Measuring Mission (TRMM) data set between the 35N and 35S latitudes. This product is a 3 hour average which the pixel values are displayed in millimeters per hour. For this project this data was downloaded at 0z for every date which the selected storms interacted with the study area.
  • 5. 4 In order to determine qualifying storms for the project, an El Nino Southern Oscillation Severity index was required to select which years were appropriate to examine. Figure 2 was used to determine ideal years to obtain data for as a strong El Nino (peak red value), strong La Nina (peak blue value), and normal condition year was required for comparison. Due to the limitation of TRMM data that it is only valid from 1998 to 2014, the seasons 2003 and 2010 were identified for strong El Nino conditions; the seasons 2002 and 2004 were identified for “normal” conditions; and the seasons 2000, 2008, and 2011 were identified for strong La Nina conditions. Of these seasons, two storms fulfilled the study area and tropical cyclone strength requirements during El Nino conditions, two storms fulfilled the requirements for “normal” conditions, and six storms fulfilled the requirements during La Nina conditions. In the El Nino years, Joel, the only storm of the 2010 season began as a subtropical storm, so the season examined for El Nino was 2003. For the “normal” conditions, 2002 was devoid of storms fulfilling the requirements, and thus the 2004 season was chosen. Also of note, the 2004 storm Figure 2 shows El Nino Southern Oscillation values rated by sea surface temperature anomaly with areas classified in red as El Nino years and blue as La Nina years.
  • 6. 5 Gafilo was the strongest storm of record to form in the Indian Ocean. As for the strong La Nina conditions, the 2000 season was chosen to represent because of its intensity of conditions and its frequency of storms. Figures 3 and 4 were also use to predict the frequency and strength of the storms in relation to El Nino and La Nina conditions. Note that in Figure 3 which depicts the sea surface temperature anomaly during El Nino conditions that the warm water temperatures is limited to the upper part of the study area. In contrast, in Figure 4 the warm waters are shown surrounding much of the island of Madagascar which suggests that the waters the storms enter before crossing Madagascar and entering into the study area provides the storms with energy to pass over the orographic landscape and retain Tropical Cyclone strength. Figure 4 shows global sea surface temperatures as opposed to the "normal" conditions during an El Nino year with an emphasis on the East Indian and West Pacific Oceans. Figure 4 shows global sea surface temperatures as opposed to the "normal" conditions during a La Nina year with an emphasis on the East Indian and West Pacific Oceans.
  • 7. 6 Table 1 shows the six storms examined with their dates of relevance to the study area in both Y-M-D, as well as Julian date. Also included is the duration in days of each storms activity in the study area. This was calculated through IBTrACS data. Name Year Month Day Hour (UTM) Julian Duration (days) Start Astride 1999 12 31 12 365 4 Stop 2000 1 3 6 3 Start Leon-Eline 2000 2 17 6 48 6 Stop 2000 2 22 12 53 Start Hudah 2000 4 2 12 92 7 Stop 2000 4 8 12 98 Start Japhet 2003 2 24 12 55 8 Stop 2003 3 3 0 62 Start Cela 2003 12 9 12 343 11 Stop 2003 12 19 18 353 Start Gafilo 2004 3 6 18 65 8 Stop 2004 3 13 6 72 Table 1 describes the chosen storms with the interaction with the study area. It includes the start time and the end time of the storms track within the study area. Also included is the date in both Year-Month-Day, as well as the Julian calendar. The time listed is the 3-hour product when the storm track intersected with the study area. The duration is calculated through the difference in the Julian days as well as the addition of the start day. This was used to select the TRMM data dates to download at 0z for these dates. The addition of the extra day regardless of the hour of interaction should make the hour insignificant in order to maintain uniform data. In order to obtain the dates, IBTrACS data was loaded into ArcMap and the highlighted storm tracks were selected out of the total. Next, the study area was defined as a polygon and a selection by location was performed to extract the beginning and end of the tracks involvement with the study area. These dates were then exported and compiled into Table 1. The final results of the track selection lines can be examined in Figure 1.
  • 8. 7 Methods: The first step is to use the ENSO Index to decide which seasons are appropriate for displaying the peak values of El Nino, La Nina, and “normal”. More information about this process was discussed in the Data section and refer any questions there. The selected storms which attain Tropical Cyclone Strength during their lifetime and cross over the island nation of Madagascar during the selected seasons of 2000, 2002, 2003, 2004, 2008, 2010, and 2011 are listed below. Of these the seasons and storms that were chosen include 2000, 2003, and 2004: Season Storms ENSO Category 2000 Astride, Leon-Eline, Hudah La Nina 2002 N/A “Normal” 2003 Japhet El Nino 2004 Cela, Gafilo “Normal” 2008 Ivan, Jokwe La Nina 2010 Joel El Nino 2011 Giovanna La Nina Table 2 shows the seasons and storms that correspond with the peak ENSO condition diplayed in the table. The table is also color co-ordinated based on the ENSO category. Of these seasons, 2000, 2003, and 2004 were chosen to represent their categories. The next step involves loading the TRMM data into ENVI for raster manipulation. One of the biggest challenges for this project is the sheer extent of the data temporally and the numerous factors that can affect the storm strength. Of the selected storms, TRMM data from every date as determined by Table 1 at 0z was downloaded and organized by season and storm. Next the raw data was preprocessed. The preprocessing of the data is one of the most time consuming factors and the TRMM data can be difficult to load properly. However, all analysis is useless if the data is not properly displayed as the project compares the rainfall to the study area, but the locations do not line up. From NASA there is a detailed list to properly load the data:
  • 9. 8 TRMM L-3 HDF:| (GES DISC; NASA) 1. From the command line, type in, ENVI, 2. Click File|Open External File|Generic Format|HDF. Select a file and click OK 3. Select both precipitation and relativeError (e.g., 3B43.20100201.7.HDF) and click OK. Select BSQ in HDF Data Set Storage Order 4. Select Basic Tools|Rotate/Flip Data. Select Data Set #1 (precipitation) and click OK. Type in -90 degrees and select File. 5. Right click the new file and select Edit Header. Click Edit Attributes|Map Info. Fill in the table with Image X = 1, Image Y = 1; -180.0 and 50.0 for Map Coordinate of the Tie Point. Put 0.25 degree for Pixel Size of both X and Y. Click OK. 6. Pick the parameter you just added the map info to and click Load Band. Once the TRMM data has been properly loaded, a spatial subset should be performed around the study area in order to focus on the location. Afterwards, for each band perform a Color Slice in ENVI from the table of contents and remove all the current color slices. Add a new slice and define the minimum at 5 and leave the maximum. This creates a selection of only the pixels which are valued at 5 millimeter per hour precipitation or greater which can then be exported to a shapefile and loaded into ArcMap. This can be measured in respect to the study area to determine the amount of space in the study area which is valued at Tropical Cyclone strength. This can be quantified and analysis can be run to compare the results to the ENSO severity index to determine correlation between the two.
  • 10. 9 Results: The results of the methods are compiled in Figures 5, 6, and 7 and represent the 2004 “normal” conditions, the 2003 El Nino conditions, and the 2000 La Nina conditions respectively. Figure 5 which represents the “normal” conditions shows the two storms Cela and Gafilo along with their tracks and the shapefiles which represent their pixel of precipitation which exceed 5 millimeters per hour. As you can see from the figure, both storms registered a fairly large amount of pixels on the maps and also followed similar paths coming from the northeast of the image and crossing the island, and then proceeding to rotate back across Madagascar. The data and the track that are green correspond to Gafilo, and the data and track that are red correspond to Cela. Gafilo was the strongest recorded storm in the Southwest Indian Ocean. Figure 6 shows strong El Nino conditions and is paired with the track and precipitation data of Japhet in green as well as the track of Joel in blue with crosses. The precipitation data for Joel was not run due to its origin as subtropical, but since the El Nino season only included one Figure 5 shows the Tropical Cyclone Tracks for the 2000 season which is associated with strong La Nina conditions. Figure 7 shows the Tropical Cyclone Tracks for the 2003 season which is associated with strong El Nino conditions. Joel track is included only to show track. Data associated with the Japhet is also included. Figure 6 shows the Tropical Cyclone Tracks for the 2004 season which is associated with "normal" conditions. Precipitation data for Cela and Gafilo is represented as well with red and green respectively.
  • 11. 10 Tropical Cyclone, the Joel track was added for comparison. Based on these two tracks, El Nino years also tend to show a similar pattern moving from around the southern part of the channel across into the mainland. Also, neither storm directly crossed over Madagascar. The origin of Japhet is also a point of interest as it is the only Tropical Cyclone in the study which originated within the channel itself. Despite all of its precipitation data being present in the image and its lack of crossing the mountainous terrain of Madagascar, very few pixels are highlighted as Tropical Cyclone strength in the channel. This could be due to the TRMM 3-hourly data recorded which was then averaged to obtain millimeters per hour, but overall, Japhet was one of the weakest storms recorded through the study. Figure 7 shows the 2000 season which was used to represent the La Nina conditions. The tracks of the three storms Astride, Leon-Eline, and Hudah are displayed on the figure. All three seem to have a uniform pattern of crossing Madagascar on the northern half of the island, the dipping south upon channel reentry and then moving onto the mainland. Precipitation data of Leone- Eline, the strongest storm in the channel of this study, is displayed in Figure 8. This is the strongest storm due to much of Gafilo TRMM values being recorded over land. Figure 8 shows the precipitation data as the Tropical Cyclone moves across the Mozambique Channel. Feb 19 and 22 produced the most catagorized rainfall.
  • 12. 11 Conclusions: In conclusion, based on the years sampled and the method of quantifying the storms, there does seem to be a connection between occurrence and strength of Tropical Cyclones which cross into the Mozambique Channel and the El Nino Southern Oscillation, specifically, years of low ENSO (La Nina) tends to have very active Tropical Cyclone seasons with the strongest storm in the channel as well as the most storms in the sampling years at six while the other two conditions created two storms each. Also in support of this is the strong El Nino conditions lack of creating a storm which crossed the island of Madagascar with one originating in the channel and the other forming as a subtropical storm. Also, during “normal” conditions, while there were two storms formed, they both occurred in the same year while the other values recorded no storms. And finally, Gafilo was such a strong storm that the precipitation display could easily be linked to other factors external of the study area. Discussion: Some of the limitations of the data include the inability to isolate the El Nino Southern Oscillation as the sole factor in the storm creation. Other teleconnections such as the Madden- Julian Oscillation (MJO) may have had an impact on the storms. Further research should focus on the sea surface temperatures of the individual storms in the region of formation and follow through to their entry into the channel. The limitation of TRMM not being reliable over land would require land precipitation gauges to be incorporated into the data, or the selection of another form of precipitation data. The inability to create an extensive profile due to the temporal range of TRMM is also a limiting factor. Another factor is the rating of when the storms in
  • 13. 12 question were recorded as Tropical Cyclones. All storms selected were rated at Tropical Cyclones during their lifetimes, but may not have been even before beginning to cross Madagascar. Due to the large amounts of images to be processed, only selecting the 0z time of the 3-hour TRMM images may have removed higher value images from the test pool, and so a more regular set of images may be required to fully understand the data. Another factor was the size of the storms, since most were too small to see the eyewall and core cloud formation for measurement and tracking purposes. Another question raised by this project was if there seems to be uniform path the Tropical Cyclones seem to follow as they cross the channel as determined by their ENSO category as the current tracks seems to suggest. And finally, by transforming the color slice of the raster directly into a shapefile before recording the values, it was not possible to easily ascertain the number of pixels valued at greater than 5 in the study area. This limited the amount of quantitative analysis that was able to be performed on the results. This project was a very large scope in order to examine a general climatologic trend, but for more accurate and definitive result, more intimate studies of each storm in question need to be performed in order to understand the impact of external factors on the final data.
  • 14. 13 References: El Niño and La Niña Years and Intensities. (n.d.). Retrieved December 4, 2015, from http://ggweather.com/enso/oni.htm ENSO Myth Number 4 - The Variations in the East Pacific and the East Indian-West Pacific Sea Surface Temperatures Counteract One Another. (2013, March 27). Retrieved December 4, 2015, from http://wattsupwiththat.com/2013/03/27/enso-myth-number-4-the-variations-in-the-east- pacific-and-the-east-indian-west-pacific-sea-surface-temperatures-counteract-one-another/ How can I import TRMM data in ENVI? (n.d.). Retrieved December 11, 2015, from http://disc.sci.gsfc.nasa.gov/recipes/?q=faq/How-can-I-import-TRMM-data-ENVI Index of TRMM Data. (n.d.). Retrieved from ftp://disc2.nascom.nasa.gov/data/s4pa/TRMM_L3/TRMM_3B42/ Matyas, C. (n.d.). International Journal of Climatology Volume 35, Issue 3, Article first published online: 7 APR 2014. Retrieved December 4, 2015, from http://onlinelibrary.wiley.com/doi/10.1002/joc.3985/pdf TRMM Tropical Cyclones - SurfaceRain Tutorial. (n.d.). Retrieved December 11, 2015, from http://www.nrlmry.navy.mil/sat_training/tropical_cyclones/trmm/surfacerain/ Tropical Rainfall Measurement Mission. (n.d.). Retrieved December 11, 2015, from http://trmm.gsfc.nasa.gov/ Vitart, F., Anderson, D., & Stockdale, T. (2003, June 19). 3932 V OLUME 16 JOURNAL OF CLIMATE q 2003 American Meteorological Society Seasonal Forecasting of Tropical Cyclone Landfall over Mozambique. Retrieved December 4, 2015, from http://journals.ametsoc.org/doi/pdf/10.1175/1520-0442(2003)0162.0.CO;2