GIS&RS news - Pacific GIS and RS Council – Pacific GIS ... · transmission of the dengue virus,...

December, 2003 Pacific GIS&RS NEWS

GIS&RS newsPacific Islands The Newsletter of the

GIS/Remote SensingUsers in the Pacific

Issue 3/2003December 2003

Highlights

o IKONOS MS Images for Fiji PIne Limitedo IKONOS MS for Fiji Forestryo GIS for Utilities in Cook and Solomon Islandso GIS&RS for Fishery Departments

Highlights

o GIS for Fiji's Health Sectoro GIS in Vanuatuo The Annual GIS&RS Conferenceo SOPAC-EU Project Updates

Time has flown. We have been producing the GIS&RSnewsletter for over ten years. The newsletter startedin 1993 where MSD-Forestry (Management ServicesDivision, the forestry inventory section) was assistedby SOPAC in jointly editing and printing the newsletter.See picture of first newsletter.In the early nineties, MSD-Forestry was the only entityin Fiji and other Pacific Island Countries analysingdigital satellite images and together with the NativeLand Truth Board the only practical users of GIS. Itwas essential to have regular monthly meetings todiscuss upcoming issues such as different projections,formats and data standards with other organisationswho were investigating GIS establishment. This wasthe start of the GIS and Remote Sensing User Group,which is now an established forum. In the first years of the newsletter there was a needto document and publish these monthly meetings toensure that all stakeholders were informed, datasharing promoted and duplication of effort avoided. Inaddition to reports of the meetings, the newsletterincluded articles covering new applications in the fieldsof GIS, RS and GPS and also provided informationabout availability of new data and new software andhardware products. Over the years the “Fiji User GroupGIS and RS News” evolved into the “Pacific IslandsGIS&RS News”. The Fiji User Group monthly meetingreports were discontinued in the newsletter and werepublished via the GIS-PACNet mailing list.The newsletter still does not need any advertisementsto cover the printing and distribution, which is done bySOPAC as most members of the editorial team workon the newsletter in their free time. The newsletter isdistributed in nearly all Pacific Island Countries witharticles contributed from many of the countries.However, this is still a shortage of articles andconsiderable time is spent communicating with

potential contributors and demonstrating the value ofdocumenting their activities and outcomes.The editorial team changed several times over theyears. Wolf Forstreuter is still looking after thenewsletter since day one when it started with him andLes Allinson. Lala Bukarau, who joined at a very earlystage to ensure a more professional editing, is stillpart of the team. To involve another regionalorganisation, training GIS and Remote Sensing, theUniversity of the South Pacific became part of theeditorial team. First James Britton gave the newslettera new outlook and today, Conway Pene is contributinghis ideas. We look forward to an increase incontribution to enable more newsletters beingpublished per year.

Wolf

10 10 10 10 10 YYYYYearearearearears ofs ofs ofs ofs of GIS&RS GIS&RS GIS&RS GIS&RS GIS&RSNewsletter ....Newsletter ....Newsletter ....Newsletter ....Newsletter ....

ISSN 1562-4250

2 Pacific GIS&RS NEWS December, 2003

First Images of OrbView-3pan-sharpened image of Bagdad availableon OrbImage's web site

Content

GIS customisation training 3GIS application in the health sector 4GIS and RS for Fisheries Departments in Pacific Island Countries 6Application of Multi-Spectral IKONOS Image Data at Fiji Pine 8Is GIS important? The Case of Vanuatu GIS Core Group 10Sixth Annual User Conference 13Use of GIS in the Power Utility (SIEA) for Analysis of Customer Billing System in Solomon Islands 14Cost Savings through the use of GIS at Fiji Pine 17Image Enhancement of Multi-Spectral IKONOS Image Data for Forest Stratificationb 17SOPAC-EU Project Updates 19

December, 2003 Pacific GIS&RS NEWS 3

GIS Customisation TrainingBen Parakoti

Department of Water WorksCook Islands

The Department of Water Works, Rarotonga GISCustomisation workshop commenced on the 19th ofNovember and ended on the 19th of December.The Department of Water Works has in place a GISsystem to manage its data of the water reticulationsystem on Rarotonga. The data are stored inAutoCAD, MapInfo, Access and Excel. The workshopis designed to expand the storage and managementof data.Discussions were held during the workshop to get thebest possible system for the department. This beganwith the physical pipe as being the smallest informationunit.The pipes installed were at 5.7m lengths ofunplasticised Polyvinyl Chloride (uPVC) and 6.0mlengths of Heavy Duty Galvanised Iron (HDGI). Usingthe join from pipe to pipe would create too manyelements and with the increasing number of joins,makes it impossible to identify these joins. This alsohas the disadvantage on the different materials linkedto one unit pipe. We came to conclusion that a pipejoin is from a valve or a node.A pipe can be of any quantity of lengths depending onthe joins at both ends. A node is at a change point orcross connection point and located on a tee. A valveis a sluice valve, gate valve or a service valve. Ahydrant is identified as a valve and in other situations itis considered as an end valve or end point. We alsohave situations that we have additional nodes betweenthe start valves or between nodes or end valve.With this situation, the programming part waschallenging, as a valve has to be connected to a pipeor a node. The sums of four valves or pipes areconnected to a node, especially at cross connectionpoints or feeding points. A node cannot be connectedto a node but to a pipe or valve.A household connection is a tapping band, which isnot considered as a node. The pipe from the tappingband to the house is considered as a pipe. Both thetapping band and the pipe to the house is recognisedas consumer or customer. All consumers orcustomers are connected on a pipe and not on a nodeor valve.The simulation program WaterFlow was written inMapBasic because it is more stable than AccessBasic. The program selects a pipe, which isconnected to a valve and sets the switch to “ON”. The

GIS Customisation Workshopprogram also selects the valves connected to pipes,nodes connected to pipes, nodes connected to valves,and identify the start valve, end valve or start node orend node.Then the program WaterFlow creates a thematic map,where all the pipes that have water flowing throughthem coloured in blue and all the pipes that do not havewater flowing through them coloured in red. Thisprogram helps us, for example in the figure below. Ifthere is a break in the line, or if there is a newconnection to be installed on the line (pipe 82), valves100 and 101 must be turned off. From this programwe can see which line will be affected and also thecustomers that will be affected as well.

DevelopmentsGIS databasing is used in development works carriedout in the Department of Water Works. The assistanceof these software’s provides a much efficient andeconomic system within data collection and storage.This also assists the department in managing its datafor past records and future developments.


GIS Applications in the Health SectorViliame Vuiyanuca & Salote Vosaicake

(University of the South Pacific)

IntroductionThe establishment and application of GeographicalInformation System within the health sector has beenrelatively slow in the Pacific. The technology is stillbeing developed providing a chance for small PacificIsland Countries to realise the importance andpotential of using spatial datasets in GIS. In Fiji, theamount of data within the Ministry of Health isenormous, and could be better utilised by presentingthem visually within the application of a GIS. There isa need to create a pilot project with respect to theapplications of GIS in the health sector to demonstratethe importance of this technology to policy makersinvolved in the decision making process.This study developed a spatial database of health-related information from the Central Division of Fiji.Though there is a lot of useful and important datawithin the Department of Health, not much of it is geo-referenced. A GIS interface was created for theavailable data to provide the end user with a set oftools for: common queries; visualising the spatialdistribution of certain public health indicators and;provide the foundation for future development intostudies of disease outbreak and distribution patterns.

Objectives:The aim of the project is not only to display datasetsvisually but more importantly to explore the potentialof GIS in the following areas:o Mapping the boundaries of nursing zones, medical

areas and sub-divisions that fall within the CentralDivision

o Location of health centerso Number of medical personnelo Location of places and population of each nursing

zoneo Population density and area in sq km for each

nursing zoneso Location of the incidence of dengue hemorrhagic

fever in the Suva- Nausori corridoro Analysis and understanding of spatial epidemiology,

in this case dengue fever focusing on theassociation between the spatial patterns of thedengue epidemic in regards to environmental andcultural factors.

Project OutputWithin the Central Division, there are approximately23 Medical Areas, 71 Nursing Zones, and 46 HealthCentres excluding village dispensaries. At the end ofthe project we managed to produce a complete datasetof the Ministry of Health’s Central Division MedicalBoundaries which include the medical areas, nursingzones, places (schools, settlements, and villages),population distribution, area of nursing zones in sq km,and the epidemic distribution map showing Dengue

GIS Applications in the Health Sector

Figure 1: Using SQL Query to select a Nursing Zone within MapInfo

December, 2003 Pacific GIS&RS NEWS 5Figure 2: Dengue Fever Reported Cases According to home address

GIS Applications in the Health Sector

Figure 3: Map Overlay of Medical Areas, NursingZones, Places, Health Centers and Roads

(DHF) reported cases at the Colonial War MemorialHospital (CWMH) between the months of January toApril, 2003.

Overview of GIS Analysis on the Incidence ofEpidemicsCase Study: Dengue Fever Reported Casesalong the Suva-Nausori corridorEpidemiology is the study of the frequency ofoccurrence, pattern and distribution of diseases in apopulation. GIS can be used with significant impactin this field.The incidence of communicable diseases in Fiji is highbecause of its tropical climate. Fiji’s geographicalfeatures favour the growth and survival of disease-carrying agents called vectors. Medical literature hasdocumented outbreaks of dengue hemorrhagic fever(DHF) in the last few decades with notable fatalities.The interaction between human beings, the host (onethat harbours the disease), and the mosquito, Aedesaegypti, the vector, is influenced by favourableconditions in the environment and by human activities.Therefore, geography plays a significant role in thetransmission of the dengue virus, the infective agent.In addition, being a developing country, socio-economic factors also influence the epidemiology ofdengue.Information collected during dengue outbreaks on thepopulation at risk, environmental conditions that favourits transmission, the pattern and distribution of thepopulation being infected, accessibility and availabilityof health care can be used to formulate strategies toprevent future outbreaks and to reduce morbidity andmortality. It is in these areas where interactive GIS

can be beneficial for health planning. To date, PICsincluding Fiji have not utilised GIS in this area of study.Last year, Fiji’s Ministry of Health launched its PatientInformation System (PATIS) where each person in thepopulation is issued a National Health Number (NHN).Each time a person uses the health system, it isrecorded in PATIS. The information derived from PATISwill help the Ministry of Health in managing itsresources. Linking interactive GIS not only forcommunicable diseases but for all diseases will assistthe Ministry of Health in managing its limited resourceseffectively and efficiently in the hope of deliveringquality health services.


IntroductionSeveral remote sensing (RS) techniques and datasources have been developed in recent years tomonitor our oceans and identify areas where fishconcentration is expected. Continued overfishing oftuna stock in the Atlantic has depleted resources andthe concern is the international fishing industry willdeplete tuna stocks here in the Pacific too. Inattempting to monitor fisheries resources, PacificIsland Countries have to compete against fishingvessels equipped with modern equipment. Eachpurse seining fishing vessel may invest in equipmentworth several million dollars however, not even one ofthe estimated 50 to 100 boats belongs to a PacificIsland Country. The high running cost between $US8,000 to 15,000 per day forces a greater perfection intuna school identification to decrease the number ofdays spent searching for schools. It is is expectedthat purse seine fishing vessels are equipped withhardware, software and data to monitor and trackresources with the cost easily paid off by reducingtime on the ocean by less than a day. The converseis true for fisheries departments in PICs where notone is equipped with remote sensing technology andknowledge to monitor their respective EEZs.The ocean is the resource of Pacific Island Countries,unexpectedly tuna prices are decreasing with declining

GIS and Remote Sensing forFisheries Departments in Pacific

Island CountriesWolf Forstreuter

SOPAC

GIS and Remote Sensing for Fisheries Departments in PICs

Figure 1: A pure seine fishing vessel. These shipsrequire a large investment and have very high runningcosts. Helicopter and other equipment help to reducethe time these vessels search for tuna.

Figure 3: ENVISAT is a new satellite which helps tomonitor the worlds oceans. Equipped with 10 differentsensors it will provide new quality of data which addsto the data available from OrbView-2, Terra, NOAAand others.

Figure 2: Pure seine fishing vessel catch a completeschool of tuna unlike local near shore fishing vesselswhich just get the tuna on the surface. This type offishing decresed the tuna stock in the Pacific and willforce the fishery departments to monitor the resource.

stocks. Besides the off shore fisheries, alternativeutilisations of the ocean could be assisted by GIS andRS applications.

Available DataPhytoplankton normally is the start of the ocean foodchain where tuna will be found soon. Areas of highphytoplankton concentration can be identified throughcolour analysis of hyper-spectral image data. Hyper-spectral sensors such as SeaWIFS on OrbView-2satellite or MODIS on Terra satellite or MERIS onENVISAT record the surface reflection in many butnarrow spectral bands, they have for example severalbands for the blue part of the spectrum unlike the“normal” land use satellites which have one only bandfor blue, one for green, etc.


This is described in Newsletter 2002-01, where GISobserves the amount of shells within a lagoon to avoiddiseases. Maps showing farm boundaries,bathymetry and image information can be produced

to assist farmers in the management of theirrespective farms.

Clam FarmingClam farming becomes increasing important as anadditional income for island communities. GISinformation such as current and bathymetry assistsin identifying areas where clams can be located tominismise spawn drifting out to deep water.

Milk Fish FarmingMilkfish can be an income on the local market, currentlymost of the production is sold as bait for long liners. In


Figure 4: Pearl shells in hanging baskets, the waypearl farming is currently carried out in Abaiang, Kiribati.Later, the single shells will be hanging down on lines.

Figure 5: Clam farming can provide additional incometo local communities. Photo from the Internet, inPacific Island Countries clam farming is carried out inshallow water where no diving equipment is required.

Another indicator for tuna catch is the sea surfacetemperature. Data can be recorded by sensors suchas AVHRR on board NOAA satellite, MODIS on boardof Terra Satellite and AATSR on board of ENVISAT.

Monitoring the Off Shore FisheryIt is important to monitor the EEZ of each country toguide their own fishing fleets even if there are onlylong liners or small boats available. There is alsoanother reason: in areas where is a high possibility oftuna foreign fishing vessels are guided by moderntechnology will be concentrated. However, asmentioned before currently there is no fishingdepartment equipped with necessary software andskilled personnel to carry out sophisticated RSanalysis.SOPAC plans to in conjunction with IRD in Noumea,New Caledonia to send pre-processed image datafrom IRD to Cook Islands . If this test turns out to besuccessful other countries will be supplied as well untilthe facilities are established within the fisherydepartments of the Pacific Island Countries.SOPAC plans as a first step to enhance the near shoremarine application, where GIS and RS will also play avital role. At the same time skills can be built up andmore complicated RS analysis methods beimplemented.

Near Shore FisheryNear shore fishery is exploited in every Pacific IslandCountry. This type of fishery provides income for ruralcommunities and is an essential part of Pacific life.Some countries support near shore fishery byestablishing FADs, which are floats anchored in frontof the coast to attract fish. Such FADs should bealways in clear water during incoming and outgoingtide and not to be touched by the sediments of rivermouths. Multi-temporal RS data would be of importantassistance.In some countries such as Fiji the landowning unitsextend to the reef flat and any monitoring of sustainablemanagement requires mapping of the landowningboundaries. Mapping of marine parts, which arenecessary in some areas to protect fish spawning,have similar GIS requirements.Surveys record the available fishing equipmentincluding storage facilities and the catch per village.GIS provides a synoptic overview.

Pearl FarmingPearl farming is carried out to large extent in CookIslands and is also now in Fiji, Marshall Islands andKiribati. It is expected that this industry will grow further.GIS enables the monitoring and management of farms.


Kiribati the fisheries department is managing itsproduction and will establish a monitoring system.

Sea Weed FarmingSeaweed farming is supported by the European Unionin Kiribati and could become an important income forrural communities. The product has to be dried beforeit can be transported to Tarawa where it is pressedand shipped to the Philippines for processing. GIS willincorporate, bathymetry and current information toidentify suitable farm areas. Mapping of land owningunits on the island is required to avoid conflicts duringthe onshore drying process.

Coral and Aquarium Fish TradeIn Fiji and Christmas Islands aquarium fish tradedeveloped into a sustainable industry. The customersin North America and Europe require the productcertified as coming from a sustainably managed area.The same applies to coral trade where small parts ofcoral are transported alive to the same regions. Acertification requires a harvest monitoring system,which must be based on detailed maps, inventories

and recording of harvest. The maps will be producedfrom satellite imagery and bathymetric data. Therecording of harvest must take the amount and thelocation into account and will be based on GIS.

Sea Cucumber FarmingIn several countries sea cucumber is a lucrativebusiness and in some cases the fisheries departmentassists this type of farming. GIS will help to monitorboth the input of spawn and the harvest.

SummaryWith the exception of Cook Islands no fisheriesdepartment of Pacific Island Countries have trainedGIS personal. At the same time tuna extraction out ofthe Pacific drives towards the extinction of the resourceas utilising high-tech equipment is not affordable byPacific Island Countries.There are other applications for management of marineresources aside from offshore ocean monitoring. GISand remote sensing can be used for the initialmanagement of nearshore resources. Oncedeveloped and users trained, more complicatedremote sensing analyses such as ocean colour andtemperature can then be applied. Once GIS is inplace and sustainable, it will be possible to upgradethe skills of the operators to also monitor the offshoreoceans. An indicator of a sustainable activity is theregular production of output.


Figure 6: Seaweed farming is an importnat sourcefor the chemical industry and provides income for localcommunities. Picture above shows the planiting stagein shallow water and the lower picture shows the productready for pressing and subsequent shipping. Picturesprovided by the Seaweed Project

IntroductionThe SOPAC EU project will purchase high-resolutionsatellite image data for each of the 8 countries involved.Areas where the Project would focus its efforts wasselected by Stakeholders. The imagery will cover thesespecific areas. For Fiji the area selected for the Projectare a) along the coral coast in southern Viti Levubetween Pacific Harbour and Sigatoka b) the Navuawater catchment and c) the Sigatoka water catchment.Fiji Pine Limited (FPL) is one of the many stakeholderswith a few Stands existing on the periphery and withinthe Sigatoka catchment area. They are also the firststakeholder to utilise the image data.Multi-spectral IKONOS image data from differentspace borne image data sources was selectedbecause:o The spatial resolution of 4m is sufficient for

Application of Multi-Spectral IKONOSImage Data at Fiji Pine

Wolf Forstreuter, SOPACIsireli Buwawa, Fiji Pine Limited


who uses only a small subset of the complete datacoverage, will always have more precise referencedata for this particular area and over time morereference data will be mapped using DGPS survey.The lesson learned is users must be able to do theirown image rectification. Therefore the original data sethas to be supplied rather than handing out the GISbackdrop as three layer GeoTIFF only. There are threemore arguments, why the original image data set hasto be handed out 1) a user such as FPL should beable to to select the bands required 2) carry out theirown image enhancement and 3) an imprecise imagecorrection and mosaicing can change the radiometricimage content.

New Image InterpretationAfter a more precise image correction two imageswere produced a) a natural colour image with thechannel combination red, green, blue and b) a falsecolour image with the channel combination nearinfrared, red, green. For both images a simple contraststretch was performed optimising forest – non-forestdifferences as well as best contrast between pinestands, gallery forest and bush vegetation. Then theimage was converted from 16 bit to 8 bit by usingERDAS viewer to IMG routine and exported to GeoTIFF.The image interpretation showed that:o False colour infrared image clearly distinguishes

between bush vegetation, gallery forest and pineforest, see Figure 1. It is essential for vegetationmapping to have the false colour imagecombination.

o Satellite image data contain more spectralinformation than the aerial photographs which

Application of Multi-Spectral IKONOS Image Data at Fiji Pine

Figure 1: The left picture shows a natural colour image combination using the red, green and blue IKONOSchannel. The right picture shows a false colour infrared image combination using the near infrared, red andgreen channel of the IKONOS image. Pine plantations are shown in dark red (1), gallery forest are shown inintensive red (2) and bush vegetation in light red (3). The spectral content of the natural colour image does notallow this separation of vegetation. Field visits will increase the number of classes which can be separated.

1

23

2

thematic mapping at 1:10,000 scale, which isthe mapping scale of most small Pacific Islandstates and the operational scale for utilities andorganisations such as Fiji Pine.

o The four image bands allow analysis of waterbody where the blue channel is required and atthe same time vegetation analysis where the nearinfra-red channel is of importance.

The first image interpretation at FPL showed that imagedata are extremely useful and the near-infrared bandprovides information which cannot be obtained frompanchromatic aerial photographs.

Image PreparationIKONOS image data is delivered as a geo-rectifiedproduct, however, from past experience in the Pacificmost image data have to be re-rectified. The extentsof the imagery are defined using global projectionswhich when compared to local Pacific Island grids thereis often a shift due to projection transformation. Inaddition, the image data has then to be converted from16-bit to 8-bit data format for software products suchas Adobe Photoshop and MapInfo to display the imagescorrectly. The re-rectification and conversion to 8 bitMapInfo image backdrop in natural colour wasperformed as a training course at SOPAC.Image display at FPL revealed that the re-rectifiedimagery was not sufficient for the scale they wereworking at and a more precise geometric correctionwas carried out using ground control points mappedin the area.This does not mean that the image rectification atSOPAC has to be improved to the standard acceptableby all stakeholders. A stakeholder such as Fiji Pine,

1 0 Pacific GIS&RS NEWS December, 2003

do not cover the infrared spectrum. This will bean important argument when FPL managementwill discuss the issue of purchasing wither newaerial photographs high-resolution or image data.

o Image data are up-to-date and the un-stockedand un-plantable areas can be mapped easily.The aerial photographs were recorded more than5 years ago and do not show the currentsituation.

o Digitising from satellite image data is faster asthe number of cases the operator has to decidenon vegetation and vegetation are less. Oneplanning unit could be done within half a day.

o Stocking of mature plantation areas is clearlyvisible.

Summary and OutlookMulti-spectral IKONOS satellite images proved to bevery useful as the infrared spectrum provides additionalinformation to separate vegetation types required formonitoring the plantations and the interpretation isfaster when using satellite image data instead of aerialphotographs.The investigation of satellite image data utilisation isongoing. Also planned are field visits to set upinterpretation keys of different stocking densitiesgrowing over different forest undergrowth. Theinterpretation keys should identify growth over certaintime periods . using this key, areas where the survivalrate of newly planted trees is too low can also beidentified.

Application of Multi-Spectral IKONOS Image Data at Fiji Pine

For further information contact:[email protected]@[email protected]

Figure 2: IKONOS image and scanned aerial photograph as MapInfo GIS backdrop. The advantage of thenear infrared providing additional spectral information is clearly visible.

"Don’t get wet in the rain when there is an umbrellalying close to you. Use it because it was made andprovided for that purpose only. If you don’t use it, youwill be soaked.”(Old northern Pentecost Proverb as retold by LateFr.Walter Hadye LINI)In the very early stages of GIS and RS development inVanuatu, new GIS projects were funded by Donors,with each project using a different GIS package. TheseProjects were not continued, contributed largely to alack of knowledge transfer to users. An additionalproblem with projects has been a lack of long-termplans as they were driven by a specific interest byoutsiders/consultants. Duplication of effort by differentgovernment and regional organisations has added tothe problems. The use of new technologies has alsoseen a wide range of equipment purchased with littleknowledge base on troubleshooting developed. TheVanuatu GIS&RS Core group wants to develop a workplan that will tie together GIS&RS activities across allgovernment departments, NGOs and statutory bodies.

Is GIS important? The case ofVanuatu GIS Core Group

Williams P. GANILEOEU-SOPAC Intern, Vanuatu


“Make use of the technology when it is with youbecause when it is gone, you will regret it”“Is GIS important in Vanuatu: The case of the GISCore group” was the topic of the paper I prepared topresent at the GIS&RS Forum, to shed some light tothe “politics” of GIS in Vanuatu, to put it in a differenttone. I intended to share the story about our userswith you because there are few things I see that needsto be addressed now so that other country’s may notface similar problems and experiences, but developsomething more constructive as they develop their GIS/RS.If the proverb is right, why are we not using theumbrella when it is lying right next to you? Use it!

departments. The Data Agreement was seen to bethe best and solution for future GIS and RS work in thecountry.The VANRIS database has almost everything onVanuatu, from population density, land use zones,farming systems, cadastral maps, contours, attitudes,schools, roads, ocean depths etc. It is the biggestdata bank the country houses now and it will be theonly one where other GIS packages can be linked to inthe future.VANRIS/GIS Core groupIn reality, the VANRIS core group is now a shadowgroup and not functioning anymore but has beenelevated to the GIS core group. The GIS Core groupwas born from the VANRIS core group made up ofadvanced VANRIS Core users set up in 1995-2000during the Land Use Planning Project (LUPO). Otherusers are those throughout the country who are usingthe tool for first hand planning needs.The GIS core group has prioritised several key areasto be the optimum goals for GIS and RS developmentin Vanuatu. From many of these priorities, here aresome objectives that the group wishes to achieve.These are:o new policies regarding GIS projects in the countryo Data sharing policies (a more coherent one which

will be debated and mandated under certain acts ofparliament).

o How data is shared among national and internationalusers/organisations

o Advice and Directions on new GIS/RS projects.o A policy paper to be submitted to parliament as of

next year for government’s recognition of the group.During 1995-2000, the VANRIS group gained respectamong many government departments, individuals andregional organisations. Planners, economists,Directors and Director Generals and political adviserswere introduced and trained on how to use themapping database for project documentations. Locallyand internationally, LUPO became the office to adviseon development issues like can a lease can be grantedfor a particular reason or not. The office helped inalleviating poor decisions on land speculation andhelped at least for policy makers to know where andwhat types of soils, land use policy is, what populationtype is in the project site etc. Out from this project, wedeveloped six provincial atlases and a national landuse policy.However, today funds have been minimised collapsingour trainings in the provinces and also among thestakeholders. The chairmanship of the group wasreplaced and given to Lands survey and later now toGeology and Mines. Having to deal with GIS and RS

The Vanuatu GIS Core Group

What is VANRIS?To begin with, our initial GIS package is VANRIS(Vanuatu Resources Information system) built inAccess database with MapInfo tab files attached asthe mappable part of the database. During the set-upof this project, three sets of Agreements were signed.One was for an organisation or an individual to be partof the project, two, a data sharing agreement and threea short-term agreement for consultants who may wishto use VANRIS in the country. Having done this underCap 83 of the Statistic Act, it paved ways for a commondata collection and sharing policy by all government


and allowing this to happen saw a great dive in theuse of GIS and RS.VANRIS by local GIS experts so to speak is spoken ofas dead. To put this in its own perspective, the grouphas ended its life and services, meetings, dataupdates, trainings, trouble shooting, computer relatedknowledge, a rather sad thing to see to have happenedamong our own users and again another of thoseprojects that did not work for us. Though it maybe seenby many as dead, we still continue to serve thoseclients who needs it but at a very low scale.Most government departments are working in isolationof each other. There are GIS projects that no oneknows of. People travel overseas to attend GIS andRS meetings without the consent of those who areable GIS and RS users. This won’t stop me to say thatmy unit still exists and doing serious mapping. Theweaknesses I am sharing have given many of us anurge to seriously accept GIS/RS development. Underthe EDF-8 project we wish to achieve new outputsand that some of the problems we have gone throughnow will be solved. We have begun this with twotrainings so far, the Utilities and Image Rectification,which has been successfully completed by our users.Another training will be conducted sometimes inFebruary after the Vanuatu Map server is installed.In terms of bad GIS projects I can direct you toEnvironment Australia Fauna and Flora project built inon the ArcGIS. The project was done under the Forestryconservation unit and the Environment Unit. Thisproject has not been completed and we don’t knowwhere it is now.A lack of education on data uses and sharing amongus has prompted exposure to many data types. Oneof the main reasons is that GIS knowledge is limited.With just a little hands-on training, I can buy theargument that we have progressed but hesitate toacknowledge the fact that our capability in using thetool is minimal. We have not been fully exposed andutilised the tool, the simple tricks in solving problemsto the full understanding of it in theory and practical.While many users have been trained at an institutionlike USP and short trainings through SOPAC, SPC etc,with minimal knowledge of GIS, we still lack the mainthings that have to be addressed among us. To speakfor all users in Vanuatu, I can say that GIS knowledgeis just taking its place among our users but pose tosay that we have made progress but at a veryreasonable level of achievements.The problems I see are; that our users are not keen tokeep up with today’s GIS works and, a lack of what isto be done when using the tool. There are projectsthat we can do to help minimise some of the dangerswe have caused to our lands. Land use planning,utilities planning, road networks etc are interesting

areas of more GIS but we can not do it as people cannot plan and have not made themselves acquaintedto technological changes and that is GIS and RS.Although some of these projects have been developedto meet proper planning and management ofresources, we continue to face operative knowledge;something, which I can say, has been the downfall ofGIS in Vanuatu. GIS users in Vanuatu are not trainedto be GIS users. Maybe there are users but nottechnicians and this has not helped in developing GISat a standard I see as a “thorough knowledge of whatGIS and remote sensing is”.Most GIS users in Vanuatu lack technological capabilityand that this is something I wish to bring to all Users inthis forum that we need an open forum for discussionon GIS and RS. I know that SOPAC has created aGIS-Pacnet list server for GIS and RS developmentthroughput the region but we have not utilised itbecause of Internet access and costs that maybeincurred by a certain department if ever it has Internetconnection. Few have gov.vu connection to the

The Vanuatu GIS Core Group

Figure 2: 4m Multispectral IKONOS imagery overEfate has been captured for theSOPAC EU Project.


government server or direct links to TVL Vanuatu underthe com.vu addresses yet many have nothing.The basic issue of planning which we have developedthroughout 1995-2000 was VANRIS. USP, SOPAC andother regional bodies have access to VANRIS. Thisset of database is the heart of Vanuatu. I wish to askthose involved with this dataset that we are giving ourlives into your hands and that security and care shouldbe cautious in using the data. VANRIS (VanuatuResources Information system) was developed toallow us to see and make very constructive decisionsat a very low scale. With the ISM in place, Vanuatu isready and willing to move forward should there beavenues to allow us go forward for very comprehensiveGIS and Remote sensing works.ConclusionsBuilding communities with GIS, a theme which the FijiGIS and RS Interested Group Annual Conference haschosen for the 2003 Conference can be argued by mycountry's group as fundamental and necessary as webroaden our peoples through media, training of highschool teachers and conferences. Let me say thatVANRIS has a total of over 60 users in the country.With over 60 users in Vanuatu with very limitedknowledge of computing, one can always argue that itis a big leap. Thank you for listening. As you have beenmade aware of our situation, my well wishers inVanuatu would like to ask that the next GIS/RS meetingbe held in Vanuatu. I am happy to carry out thisstatement because I know how have missed so muchin the Vanuatu and I want some of you to come andpresent your work to us.Thank you blong lisen I kam long mi.

Sixth Annual GIS and RS User Conference

The Sixth Annual Pacific GIS and RS User conferencewas held from November 25-27 at the University ofthe South Pacific (USP) Marine Studies complex. Thisannual event is a highlight on the calendar for the GISand RS user community in Fiji, and is increasinglyattracting a wider audience from the Pacific region.The idea for the conference was first put forward in1998, by the Director of GIS at USP at the time, JamesBriton. The first conference, in December 1998, wasattended by a total of 70 participants. Since then,attendance has grown every year and in 2003, 180people attended over the 3 days of the conference.The theme for the 2003 conference was ‘Building

Pacific Islands GIS and RS Communities’. The themewas chosen in recognition of the increasing efforts todevelop GIS and RS in many of the smaller PacificIslands countries through collaborative efforts withmore experienced users. The theme also reflectedthe increasingly collaborative nature of many GIS andRS applications in fields that had not used GIS andRS tools technology before.The conference was opened by Dr Russell Howorth,the Deputy Director of SOPAC, after a welcome fromProf. Randolph Thaman of the Geography Departmentof USP, and Prof. Rajesh Chandra, acting ViceChancellor of USP. Dr Howorth stressed theimportance of ocean resources to Pacific Islandscountries, and outlined SOPAC’s approach to thedevelopment of these resources using GIS and RStechnology.The conference had three main areas of focus. Firstly,the presentation of updates from many of theorganisations attending allowed attendees to have aquick overview of the state of GIS and RS in Fiji andthe wider region. The updates allowed organisationsto spread information about their progress during theyear and any special success and challenges theywished to share with the wider user community.Secondly, organisations and individuals presentedspecial projects that had been developed over the year.This saw a number of presentations of pilot projectsfrom a variety of organisations and included projectsfrom USP students. These pilot projects provide agood indicator for future directions of GIS, and allowedfor discussion of the issues and challenges facing theirwider implementation. Thirdly, the conference provideda unique environment for the GIS and RS communityto meet and share ideas make contacts in the informalsetting of the conference social events. A cocktail onthe opening night, dinner on the closing night, and aday cruise on the day after the conference, gaveparticipants to relax and meet in an informal socialatmosphere.The conference was funded by sponsorship from avariety of organisations and companies, whosegenerous contributions make the event possible. Inorder to make participation available to as wide anaudience as possible, the conference does not chargeany fees. The organisers would like to thank the FijiLand Information Council; French Embassy (Suva);Telecom Fiji; Native Lands Trust Board; SOPAC; USP;GeoSystems NZ; Lukemine Enterprises; BentlySystems and Intergraph NZ. A full report on theconference is currently being compiled, and will beavailable in January 2004.For a copy of the report, please contact Conway Peneat the Geography Department at [email protected]

Sixth Annual User Conference

Coway PeneUniversity of the South Pacific


Use of GIS in the Power Utility in Solomon Islands

Introduction:This paper describes the use of GIS as a tool for theanalysis of the billing system within the SolomonIslands Electricity Authority (SIEA). The paper beginswith a brief summary of the development and progressof GIS within SIEA as tool for asset management in apower utility. The subject matter is divided into three(3) sub-topics to show the relevance of GIS and itsuse for asset management. These are as follows:

o GIS Development in Solomon Islandso Accessing Information from GISo Use of GIS for Analysis of Revenue Loss within

SIEA.GIS was first introduced to SIEA in 1996 when underthe then European Union (EU) sponsored Forum PowerUtility Training programme two of the Authority’sengineers were invited to participate in the first GISWorkshop for Power Utilities held in Nukualofa, Tongahosted by the Tongan Electricity Power Board (TEPB).As a follow up from this workshop, another GISworkshops was held in Lautoka, Fiji in April 1998.Following this workshop and the signing of theagreement with the Trimble representative in Fiji, anda subsequent request to SOPAC, staff from the formervisited SIEA in August to1998 to set up the Base Stationfor GPS surveys. In October 1998 another GISWorkshop was conducted at the SOPAC facilities inSuva, Fiji mainly to add photographic image backdroplayer to the cadastral map of Honiara. In this caseonly Honiara has been targeted, as it was the only SIEAoperational centre with aerial photos. Since then theAuthority had got two staff working full time on theproject, doing GPS surveys and collecting data for theSIEA GIS.

GIS Development in SIEAGIS was first introduced to SIEA in 1996 and 1998through Workshops sponsored by UNDP under theauspices of the Forum Secretariat’s Power UtilitiesTraining Programme which effectively came to a haltat the end of 1998.The training programs for the GIS include training onthe following:

o Database Building – Microsoft Access

o Programme Writing – MapBasico Creating Map Layers – MapInfo

After having undergone the training the GIS staff wereable to create tabular databases for the distributionsystem and customers, and link these tables inMicrosoft Access. They were also able to get thecadastral map and photographic images of Honiaraestablished in MapInfo and were able to write programsin MapBasic to create thematic maps showing thelocations of the distribution assets and also SIEAcustomers in Honiara.From 1998 till 2003 the GIS programme hadprogressed although slowly with all work beenconcentrated on Honiara. Despite the ethnic tension,which came about in the early stage of the systemdevelopment in 1999 the GPS surveys still continuedduring this difficult time and it is still progressing.Thus far the following areas have been covered, somein greater detail others still to be completed.

o Data collectiono GPS Survey datao Data Inputso Program writing

As far as data collection is concerned, about 98% ofthe data for distribution lines including transformers,poles, switches, fuses etc., and customers havealready been completed. The remaining 2%outstanding work is mainly to do with verifying andchecking the integrity of the data collected.In the area of GPS Surveys, 95% of he distributionsystem in Honiara has been completed. This includesGPS surveys of power poles, substations, switches,fuses and distribution lines such as cables, andconductors. In addition GPS surveys of 65% of thecustomers in Honiara had also been done.Most of the data collected including the GPS surveyshad been inputted into the GIS programme withprogammes written in MapBasic to create thematicmaps of the various assets such as poles, fusessubstations and customers and to show the locationof these on a map of Honiara in MapInfo.GIS is still in its developmental stage in SIEA andalthough there had been success in mapping out thedistribution system assets for Honiara there is still alot of work to be done to cover other eight operationalcentres of SIEA out in the provinces. Although therehas been interest for the GIS program shown from themanagement level that has really boosted ourenthusiasm for improving use of the GIS program itselflack of resources, mainly transport, had restrictedprogress.Following the recently held workshop at the SIEATraining Center in October 2003 funded under EU and

Use of GIS in the Power Utility (SIEA)for Analysis of Customer Billing

System in Solomon IslandsRobinson Wood, Martin Rasu,

Mahli SanauSIEA GIS Section



conducted by Dr. Wolf Forstreuter and ElizabethLomani (both technical staff of SOPAC), thedevelopment of the SIEA GIS program can only improveto a much higher level but much depends on thesupport and recognition given to this developmentprogramme in terms of resources and the benefits thatcan be derived in the area of asset management.Having participated in this workshop, the GIS Sectionof SIEA can only improve on developments to date andin the long run build a GIS that will enhance the assetmanagement capabilities of the Authority as far as thedistribution system and customers are concerned.With the use of GIS we have access at the click of abutton to information, which would otherwise takelonger to access under current SIEA operatingconditions. In addition we can save time and reduceexpenses as distribution system faults, customercomplaints and enquiries can be pinpointed moreaccurately and fault crews dispatched to the locationswithout having to search around, as its current practicewithout GIS. This is because you have informationneeded right in front of you available on the PC andcopies can be printed out for those requiring hardcopies of the information.Another benefit of GIS is that it provides informationon both spatial and non-spatial data. The GIS set upwould also assist departments concerned, mainlyengineers who often need up to date informationrelated to the distribution system.

Accessing Information from GISThis section deals with how we can access anyinformation using the GIS system. Although the SIEApower utility GIS is yet to be completed with regards toGPS surveys, and data collection and collation, workdone so far could already be used to analyze some ofthe distribution system assets and customers in theareas covered. It will take time and resources to fullycomplete the GIS system for all of SIEA’s OperatingCentres.Despite using the latest software version of theprograms required for GIS, there is need for ongoingtraining in the field of program writing or databasebuilding, such as the one recently conducted bySOPAC as SIEA do not have the expertise in this area.Currently information on distribution poles, substations,transformers, fuses, overhead conductors,underground cables and customers can be accessedfrom either the database in MS Access or from withinMapInfo. To access information from MS Access wecan go to the SIEA Main Tabular Database and clickon the relevant section.Because GIS is still in its developmental stage in SIEA,its success and impact on the area of asset

management cannot be fully realized at this stage.However, benefits of the system are already evidentfrom the information that can already be provided atthis stage.One area that would need attention and focus in thelatter stages of the development of GIS, is the trainingof SIEA staff from Departments who will be using GIShow to use the programme to better manage therespective assets and how to access informationstored on GIS. This would also assist in making surestaff have access to information to provide a betterservice to the valued customers and the public at large.

Use of GIS for Analysis of Revenue Loss withinSIEAThis third and last section of the report is an exampleof how information stored on the SIEA GIS is used toanalyze revenue losses within SIEA.One of the aims of the SIEA GIS was to assist inproviding information for identifying customers who areconnected to the SIEA distribution system network,such as quarterly energy usage averages, usagepatterns and also customers with zero readings whocould be disconnected or who are still connected butwith meters not been read. Early in 2003 it wasidentified while running a query on customers with zeroreadings that there was a substantial number ofcustomers with a zero reading. This prompted theAuthority to carry out a Customer Audit early in 2003,a task, which was never completedAnother query on the customers with a zero readingwas done recently in November 2003, which showsthat there are still a substantial number of customers(4,000 +) with zero readings. Some of thesecustomers are relocated in areas outside of the HoniaraTown Boundaries such as Foxwood, CDC andKakambona areas mostly affected by the ethnic

Figure 1: Tabular database at SIEA



tension. However, a large portion of the customerswith zero readings are also within the Honiara Townareas.This is a matter of concern and need follow up as theAuthority could be loosing revenue if some of thesemeters had been displaying zero readings for overseveral billing cycles.GIS will assist in providing thematic maps to showlocations of the customers with zero readings so thatCustomers Services Section of the Authority can checkthese out and verify the status of the readings.One of the other areas, which was identified duringrunning of the GIS queries, was the number ofcustomers with no names against installationnumbers. It was only after these were pointed outtogether with the customers with zero readings thatthe authority then set up a task force,which is responsible to carry out an auditsurvey to find customers who don’t haveright names but using consumer insteadin the system (see figure 2). As such theauthority cannot send or post any bills tothese customers and again the Authoritywas not collecting its dues from suchcustomers and hence loosing revenueIn these cases the authority is losingrevenue on the billing system. Fortunately,the customer update information wasmade to the billing system in mid 2003,which should assist in rectifying customerrecords.One of the major objectives of SIEA is toprovide efficient service for customerwhile receiving better revenue in returnthrough its billing system. This should befurther enhanced with the use of GIS

program, which we hope will create a much betterservice to everyone.

ConclusionIn conclusion it can be seen that GIS is a great tool forasset management if utilized properly. It could providesubstantial benefits for the Authority both in terms ofreduced maintenance costs and reducing revenuelosses.As one who had been working on the SIEA GIS, I feelconfident to say that GIS is a significant tool, whichenhance better servicing both within the Authority andto its valued customers. Despite setbacks andconstraints encounted, we have made it through andfeel rewarded when we saw some of the successachieved. It is of concern that the customer audit task

was halted as this would greatly assist infinding out about what customer problemsmay be out there. Lastly, I think the successof these programs depend on thecooperation of all SIEA departments whowould be users of GIS once fully developed.The Heads of Departments and/orimmediate supervisors in the distribution andcustomer services groups should visit theGIS Section and discuss with the GIS staffabout their requirement so that they can takethe necessary action to organize theirrespective work groups to get most benefitout of the GIS. This will help to provide goodworkmanship and an environment in whichwe will all enjoy and also proud of.

Figure 2: Customer location, name and installation number.

Figure 3: Display of last month consumption (red bar) and averageof last five months consumtion (green bar)


Cost Savings through the Use of GIS at Fiji Pine

The Fiji Pine's GIS has made significant progress onlysince early 2003 and are already reaping benefits interms of cartage costs. Our cartage rates aredependent on distance between the Tropik mill in Drasato the areas of operation.During the months of January and February, Fiji Pine(FPL) recorded a loss of $34,000 in cartage costs. Asfrom March FPL started calculating road distancesfrom data captured in GIS, and this provided moreaccurate distances within minutes whereby reducingthe cartage rates. As at the end of 2003 FPL recordeda savings of $292,700, resulting in a net savings of$258,700. This was possible with GIS that has providedmore accurate distances. Having realized the savings,distance determination by GIS was used extensivelyby Area managers in the preparation of our 2004budget for:

1. Road Maintenance2. Construction of New Roads3. Cartage Costs4. Preparation of Fire Breaks

At the moment FPL has started to use GIS in thepreparation of logging plans. It must be noted thatthese plans were produced previously using theconventional methods, which was never archived. FPLnow intend to store these plans in CD’S so that in thenext 15yrs when we are ready to harvest the 3rd rotationthe logging plans will be available at a press of a button.

Cost Savings through the Use of GISat Fiji Pine ltd.

Satya NandFiji Pine ltd.

Figure 1: GIS display of transport routes from thelogging areas to Drasa mil

IntroductionThe Navua Catchment was selected as a test site toapply high-resolution image data (IKONOS) fordifferent use. Fiji’s forestry department is one of thestakeholders of the SOPAC-EU project and will testthe potential of this image data for forest inventory. Animportant part of any forest inventory is the mappingof forest cover, where the exact forest area has to bedetermined. IKONOS image data are the first type ofspace borne remote sensing information allowingforest mapping at a scale of 1:10,000. During themapping process not only the boundary between forestand non-forest area has to be identified it is alsoimportant to identify and map as many different foresttypes as possible. The forest canopy has to be stratifiedinto sub-areas of higher homogeneity. When the forestmapping is completed the number of sample plots canbe calculated where inventory teams measure thestanding timber volume. Furthermore the canopy isstratified into more homogeneous sub-areas as lesssample plots are required for the same statisticalaccuracy. Therefore, the potential of forest stratificationis directly linked to the cost of the inventory.Fiji is planning to undertake the National ForestInventory and the Management Services Division(MSD) of the Forestry Department is testing potentialmulti-spectral IKONOS data to visually differentiate thevarious forest types in the image. The degree ofstratification will depend on the ability to visually interpretspectral differences within the forest canopy.Image Preparation and FieldworkThe in-office data preparation concentrated on a) bandselection, b) different image contrast stretches and c)arithmetic image enhancements such as ratios andprincipal component transformation. Then, a field teamwent on a two day field trip into the area using a Garmin12XL GPS to determine the exact location on theground which were marked on the satellite image.Therefore prints of the image were taken into the field.The area selected for the verification is located alongthe logging road of Tubarua to the riverbank oppositeto Namuamua village in Serua, see Figure 1.Observations:Site 1: This site had the highest concentration of darkred spots; the site is of high-density mahogany

Image Enhancement of Multi-SpectralIKONOS Image Data for Forest

Stratification

Wolf Forstreuter (SOPAC)Akosita Lewai (Fiji Forestry Department)


IKONOS Multi-Spectral Image Data for Forest Stratification

plantation (Galoa Plantation).Site 2: This site is of the bank of the Navua River andsome neighbouring settlements; vegetation is mainlyscattered multiple use forest (ferns, bamboo, grassand shrubs).Site 3: The site shown is mainly of scattered forestvegetation. The site is one of the old logging areasand there is some evidence of re-generation, andleftover stock is too young be logged.First Recommendationsa) Image mosaicing has to be performed withoutadjusting the contrast between the images. IKONOSimages are recorded at different days and with slightlydifferent view angles. When adjusting the contrastbetween the images the contrast is reduced betweendifferent image features at the same time. It was notpossible to see the difference between mahogany andnatural forest.b) The near-infrared band contains importantinformation for forest stratification. Mahogany wasclearly visible and could be separated from indigenousforest. It is essential to use either a band combinationblue, green, near infrared or green, red, near infrared.The natural colour combination blue, green, red didnot show the required contrast.c) IKONOS images provide usable information for thestratification of forest cover in Fiji and will allow morestratification than would be possible with panchromatic

or natural colour remotesensing data. Because it isextremely difficult to recordinfrared photography in acountry like Fiji, IKONOS datacontaining the near-infraredportion provide a costeffective solution for foreststratification at 1:10,000 scale.Further ImageEnhancementAfter the field work, differentadvanced imageenhancement techniqueswere tested together with FijiHardwood Corporation and FijiPine Limited. The followingonly describes the techniquesidentified as useful.Image MaskingThe outline of a mahoganyplantation was digitised inMapInfo and exported asShape file to ERDAS, wherethe file was converted to a

raster layer, which was used as an image mask. Usingthis mask it was possible to display only the pixels ofthe plantation area. The image stretch was notinfluenced by non-plantation areas and showed morecontrast allowing a better stratification of forestvegetation.Image RationingThrough dividing pixels of the near-infrared image bandby corresponding pixels of the red image band theinfluence of slope shadow was reduced and thecontrast between Mahogany, natural forest and un-stocked areas was enhanced at the same time. Theresult of the image ratio is a new image channel.Visualising the Ratio ImageThe normal RGB (Red, Green, Blue) image format wastransformed into IHS (Intensity, Hue, Saturation) imageformat. Hue was then replaced by the ratio image andthe IHS was transferred back into RGB. The resultallows a fast mapping of areas where mahoganydominates the forest canopy.Summary and Recommendationsa) A suitable image enhancement cannot be performed

without test areas selected in the field.b) Image enhancement for forest vegetation in Fiji

requires more advanced image enhancementthan band selection and linear contrast stretch.

c) Advanced image enhancement requires theknowledge of physical parameters of the image type

Figure 1: The image was produced using the channel combination green, redand near-infrared. The test sites are outlined in red and the GPS waypoints aredisplayed as yellow routes.


IKONOS Multi-Spectral Image Data for Forest Stratificationand basic knowledgein statistics.d) SOPAC hasto provide training totransfer thisknowledge totechnical staffworking onvegetation mappingto ensure that the fullimage content ofIKONOS images isutilised.Fiji HardwoodCorporation, Fiji Pineand Fiji ForestryDepartment still useo n - s c r e e ndelineation andsubsequent digitisingto map the differentforest types. As a nextstep SOPAC has to train

techniques transfer a computer-assisted imageanalysis.

Figure 4: This "synthetic" image was converted from RGBto IHS, then hue was replaced by a ratio infrared/red andconverted back to RGB. All Mahogany dominated partsof the canopy are shown in blue other parts are shown indiffering red tones (different part of Galoa plantation).

The EDF9 Consolidation Agreement between ForumSecretariat and SOPAC became active from the endof March. Interviews for the EDF8 Oceanographer andthe four EDF9 Advisor positions have been completed.Appointments once finalised shall be announced.

FIJIAn introductory GPS/GIS course was held March 15 -19 for National Disaster Management Office, SecurityDivision and the Fiji Red Cross.Landslide Assessment Project for the Navuacatchment is ongoing with Forestry and otherstakeholders and includes training for stakeholders,May 17 - 21 The objective is to develop a methodologywhereby stakeholders can identify actual landslidesand potential threats based on easily accessible andexisting data sources within a GIS&RS environment.Sandra Melzner, a Masters student at the Universityof Bonn will to assist in the development of a regionalmodel that could be applied to the rest of the countryReg Sanday, Resource Economist, has been assistingPartners in Community Development - Fiji in developinga model for community vulnerability analysis. PCDFis applying the model to villages in the Lau Group underan AusAID regional project implemented by the FSPIWolf Forstreuter, GIS&RS Specialist and ElizabethLomani, GIS/RS/ICT Project Intern have been assistingFiji Electricity Authority in GIS databasing andcustomisation

Figure 2: The polygon ofthe plantation boundarywas digitised in MapInfoconverted to Shape file,imported to ERDAS andconverted to a rasterlayer used to mask out allnon-plantation areas.

EDF Project UpdatesLitea Biukoto

Figure 3: After masking out all non-plantation areas,the ehancement concentrates on pixels which showforest cover only (part of Galoa plantation).


SOPAC EU Project UpdatesThe first training on image enhancement for vegetationanalysis included Fiji Pine Limited, Fiji HardwoodCorporation and Department of Forestry, April 27 –27. Fiji Hardwood and Forestry are applying thetechniques in stratifying vegetating cover to identifybroadleaf within coniferous areas. Broadleafvegetation provides stability to the soil in valleys thuswould be delineated as non-logging areasFiji Hardwood is mapping mahogany plantations wheremahogany trees have grown through the canopy andare visible on the satellite imagery. They now stratifymahogany areas into mahogany and leftover naturalforest plus other species planted using the nearinfrared band which provides additional information.Though these techniques are possible with aerialphotography, high resolution satellite data is a lot morecost effective. Fiji Hardwood has since planned to mapremaining mahogany plantations using satellite data.

KIRIBATIFranckMartin, ICT Specialist, set up the MapServer(http://map.gov.ki) at the Ministry of Fisheries andMarine Resource Development in Bairiki and followedit up with a workshop for stakeholders in uploadingdata and updating the site.

NIUEMichael Bonte, Risk Analyst and Litea Biukoto, ProjectAssistant were in Niue 23 April -04 May to assess the wave andwind impact of TC Heta on thebuilt and physical environmentand document eyewitnessaccounts of the event. Theoutcomes will be incorporatedinto the numerical modelling ofthe wave impacts on Alofi. A 3Dfly-through developed from themodel will provide a visual of theactual impacts of TC Heta andidentify potential areas affectedshould another event of thesame magnitude occur.The post TC Heta IKONOSimagery was purchased andhas been valuable for mappingout affected areasPAPUA NEW GUINEAFranck is in Port Moresby to consult with stakeholdersissues related to the MapServer and identify a possiblelocation for it

SAMOAThe data from the swathmapping survey is beinganalysed by Quan Chung, Technical Assistant

Reg represented the project at a 4-day meeting in Apiaof the Pacific Group of ACP countries in discussionwith the EU about procedures for implementing EDF9projectsReg is planning to undertake an in-country visit 24 -31 May for consultations on Project issues.

SOLOMON ISLANDSReg met with Provincial Council Stakeholders inHoniara , Malaita and New Georgia and visited theGIS centre

TONGAWhilst enroute to Samoa the MV Turagalevu detouredto Niuafo’ou aka Tin Can Island to map the bathymetryof the island. The island got its name from the uniquemail system where incoming and outgoing mail weresealed in 40-pound biscuit tins and transportedbetween the ship and island by swimmers

TUVALUFaatasi Mautama, Director of Lands, uploaded someTuvalu datasets whilst on attachment at SOPAC, tothe Tuvalu MapServer. The remaining were uploadedby Meelina and Tesimita Ailesi, GIS students at theUniversity of the South PacificThe proposed starting date for the swathmapping

survey of the Tuvalu atolls ismid- to late May

VANUATUWolf and Elizabeth conductedtraining in GIS customisationusing MapInfo, MapBasic andMicrosoft Access., March 22 -April 09Department of Lands andSurvey have expressed aninterest in Database handlingand GIS customisationFranck worked with WilliamsGanileo, Country Intern, insetting up the MapServer at theMinistry of Finance. Theuploaded data include a portionof the VANRIS datasets. <http:/

/map.vanuatu.gov.vu>Stephen Booth, Project Leader & Water ResourcesSpecialist consulted with Mines & Geology andEnvironment staff, particularly in relation to the Tagabeand Teouma catchments and a field visit to Bouffalandfill was undertaken. A reconnaissance visit wasalso undertaken to Malakula (Lakatoro) to reviewdamage to rural water supply installations followingTC Ivy.

Alofi, Niue's business district pre (left) andpost (right) TC Heta. The post Heta image isa pansharpened IKONOS image, 1mresolution

GIS&RS news - Pacific GIS and RS Council – Pacific GIS ... · transmission of the dengue virus,...

Documents

Transcript of GIS&RS news - Pacific GIS and RS Council – Pacific GIS ... · transmission of the dengue virus,...