Disaster Analysis of Kodagu District using Geomatics Technology

Disaster Analysis of Kodagu District using Geomatics Technology

Prof. Rashma Jain

HOD, Dept. of Civil Engineering Shree Devi Institute of Technology Kenjar, Mangalore

India

Sreelakshmy Madhusoodhanan, Arun S Nair, Mohammed Azmal and Niriksha Rani

Dept. of Civil Engineering Shree Devi Institute of Technology

Kenjar, Mangalore India

AbstractThis paper deals with the study of series of flashfloods followed by massive landslides that left the district of Kodagu in a devastated condition during the monsoon season of 2018. The factors generally responsible for the landslides or slope failures viz., lithology, slope, geomorphology, land use/land cover, drainage density and rainfall data are considered and each contributory feature is transformed to a thematic map by using Arc GIS version 10. Also suitable permanent restoration measures have been suggested after gaining a complete understanding of the various causative parameters regarding the calamity.

Keywords Flashflood, landslide, kodagu district, Arc GIS, landuse/ landcover, DEM.

1. INTRODUCTION

   From the social point of view, An extreme event within the earths system that results in death, injuries to the human and damages vulnerable goods is called as Disaster.

   Scientifically behind every natural disaster that occurred on the earths surface at least a single, or a multiple Geological phenomenon or earth system processes are involved in an active, continuous and in cyclic manner. These natural phenomenon or process vulnerable to human and other living being , their property as well as environment are known as Geohazards.

   Whenever a society is facing a huge loss or damage to the life / land, humans property and / or to damage to its environment due to Geohazards, then that event is called as a Natural Disaster. However all the Geohazards cannot become disasters because only few hazardous events are causing disasters. Now a days our human community have started facing disasters every frequently. This is because of our own intervention with the nature improperly though the various departmental activities and implementation without understanding such natural earth system process.

   The heavy rains and floods followed by landslide in the district of kodagu, this year has left the agricultural state in a devastating condition. The district whose economy is driven by agriculture, including coffee, pepper and paddy produced 1.16 lakh tonnes of Arabica and Robusta varieties of berries in fiscal 2017-18 with 45,000 growers producing 40 percent of Indias coffee , kodagu contributes a major quantity of aromatic beverage . The farmers in kodagu were expecting a good yield in the year 2018-19 , but due to heavy rains and floods about 60 percent of berries in the district have been destroyed , the top soil has

   been washed off in the most of the areas and the plantation lands has been left undistinguishable as of now. Also the geological survey of India has made an inventory of over 150 land slip prone spots in Kodagu alone.

   The Landslides are resulted due to variety of natural slope failure processes such as weathering, soil erosion, rainfall, high speed winds, earthquake or minor tremor, lack of vegetal cover along hill slopes and hydrostatic imbalance within the hill slopes. Rainfall, cyclic and regular geomorphic processes and their agents which include mountain building processes can cause landslides naturally. Several types of landslides and related terminologies based on the type, speed and direction of sliding materials involved are: soil creep, land/soil slip, translational slip, rock/debris slump, rock/debris fall, etc. India is vulnerable to different natural hazards due to its geodynamics and unique climate. Estimates show that about 15% of the total area is susceptible to landslides. The extensive landslides that occurred in Kodagu is an example of human induced landslides. Induced landslides are caused by the extensive weathering, slope erosion and the landslides from the hilltop which are resulted by unstable slope due to toe removal along Ghats road sections, deforestation and other anthropogenic interventions in the form of unsafe constructions along slopes, improper drainages, etc.

   1. DISTRICT PROFILE

      Kodagu, also known as Coorg, the smallest district in the state of Karnataka. It is a picturesque, hilly district located in southwestern Karnataka, on the Western Ghats of India, and is considered as one of the most beautiful hill stations of Karnataka. The land of Kodagu is blessed with exceptionally beautiful natural landscapes. The region is well suited for the cultivation of many high valued plantation crops like Coffee, black pepper and cardamom. Thus, the district is driven by a thriving rural agriculture based economy, which is supplemented by incomes from tourism.

      Fig.1, Kodagu District Map

   2. RAINFALL

      The district enjoys typical tropical climate characterized by slight to medium humidity due to proximity to coast. It is known to be quite pleasant and healthy, characterized by high humidity, heavy rainfall and cool summer. A major part of the year consists of rainy season as the monsoon period starting in June lasts till the ends of September. Even during the post monsoon months of October and November certain parts of the district receive a significant amount of rainfall. The analysis of the last 10 years data reveals that the highest rainfall (Average 3302.46 mm) has occurred in Madikeri taluk located in the Western part of the district which is thickly forested and the lowest (Average 2105.22mm) in Somvarpet, taluk which is in northern part of the district having less forest cover and adjacent to Maidan (Plain) region. Topographic influence on rainfall is clear from the spacial distribution. Therefore, the rainfall is found to go on decreasing as one proceeds from the western part of the district to the eastern part.

2. OBJECTIVES
   1. 1. Analysis of flood and landslide consequences in the region by statistical data interpretation.
      2. Study of Geomorphology of the area in order to access the hydrologic response using Geomatics technology.
      3. Identification of landslide prone zones in the district.
      4. To suggest Landslide mitigation practices in the area based on the study outcome.
3. METHODOLOGY

   The study is conducted by dividing the district taluk wise. The sites for investigation were selected using toposheets (open series toposheet nos: D43V14, D43V15 and D43V11) and on the basis of intensity of damage caused by flood and landslide (identified from statistical records

   maintained by PWD of Madikeri, Somvarpet and Virajpet Taluks).In Geomorphological study the thematic mapslike Drainage map, Lu/Lc map, Slope map, Geomorphology map, Aspect map, Hill shade map, TIN will be prepared using Arc GIS 10 and remote sensing products like satellite images and DEM data along with ground truth information. Initially georefferencing and image corrections has to be done for the toposheets and satellite images respectively. Once we obtain the boundary of the study area we need to work on the different thematic maps namely drainage map by obtaining the DEM data (clip according to the base map boundary). Then the drainage map is prepared using Strahlers method of stream ordering. The geomorphological study is carried out by quantification of geomorphological parameters.

   The relevant thematic maps for study are extracted. The relief parameters are determined (using DEM data as it has elevations in it). That is, the mutiple thematic layers will be integrated in GIS Platform to predict the hydrologic response. We have made use of ERDAS 9.2 for remote sensing and Arc Map 10.2.2 for GIS software.

4. STUDY AREA

   The study area lies in SOI toposheet nos. D43V11, D43V14, D43V15 in parts of Kodagu district, Karnataka and is bounded between latitudes 12.4240N and 12.5940N and longitudes 75.7380E and 75.8500E. The study area is located in the high precipitation zone with picturesque topography occupying the eastern and western slopes of the Western Ghats. The study area is well connected by highways and other main roads. We have selected a buffer area of 10km radius surrounding SH 37; 40km road connecting Madikeri Taluk to Somvarpet as our area of interest and have conducted all our study in this region to get a clear understanding of the topography and associated factors, the intensity of the disaster, the triggering factors or elements. Around 100m continuous stretch of state highway No.37 has been washed downstream by the flood water leaving the area completely disconnected for a few days and also making the rehabilitation process to be difficult.

   Fig.2, Base Map of Area of Interest

   A. ANALYSIS OF RAINFALL DATA OF MONSOON 2018

   During monsoons of 2018, Kodagu has experienced very heavy rainfall. The available data shows that from 1st January to 31st August 2018 (period of 8 months), cumulative rainfall in the Kodagu district is 3464 mm which is 32% more that the average annual rainfall for last

   20 years. Madikeri and Somwarpet taluk has received cumulative rainfalls of 4692 mm (32% more compared to the average annual rainfall for last 20 years) and 2701 mm (28% more compared to the average annual rainfall for last 20 years) respectively. It should be taken into consideration that at least one month of peak monsoon and three months of post monsoon period is yet to come. So, the annual rainfall figures for the year of 2018 will certainly increase from the available data.

   Analysis of Rainfall data

   Analysis of Rainfall data

    

   5000

   4500

   4000

   3500

   3000

   2500

   2000

   1500

   1000

   500

   0

   Average annual

   rainfall of last 20 years

   5000

   4500

   4000

   3500

   3000

   2500

   2000

   1500

   1000

   500

   0

   Average annual

   rainfall of last 20 years

    

   Actual cumulative

   rainfall for 2018 from 1st Jan to 31st August

   Actual cumulative

   rainfall for 2018 from 1st Jan to 31st August

    

   Fig.3, Analysis of Rainfall data till August 31st

5. RESULTS and DISCUSSION

Thematic maps generated using ArcGIS software, DEM, toposheets, satellite images and google earth data:

1. DRAINAGE NETWORK

   Fig.4.1, Drainage map of study area prepared by digitization of toposheets.

   Fig.4.2, Drainage map generated from digital elevation model

   Fig.4.3, Drainage map of study area extracted from Google earth image.

   A drainage map shows the watershed of an area, the direction and elevation that water flows from into a stream, river, or lake. The drainage distribution is dependent on complex variables such as morphology mainly relief, slope, structure, lithology, topography, climate and specific indicators provide the clues to general and specific indicator slope in the area and constitute important component of surface hydrology. After the preparation on the base map the drainage map can be prepared. With the georeferenced toposheets the drainage lines are traced and then these are updated using the satellite imagery for more accuracy. For the steam ordering we have used Strahler ordering method. There are two methods proposed by Strahler (1957) and Shreve (1966). The Strahler method was used in this study, it is the most common stream ordering method. With Strahler method we assign an order of one for all links without any tributaries and are referred to as the first order. The intersection of two first order links creates a second order link, the intersection of two second order links create a third order link, and so on.

   Here we have prepared three drainage maps by three different methods. The drainage map prepared from March 2011 DEM data has an accuracy ranging between 60% to 80% whereas the drainage map prepared by digitization of toposheet (surveyed during 1968-69 and last modified during 2006-07) has an accuracy of 80%. The third drainage map prepared from google earth depicts the present state of the drainage system but this methods accuracy is less than 60%. However the major changes during three time periods 1968, 2011 and 2019 can be identified.

   It has been observed from the three maps that many of the smaller first order streams have dried up reducing the drainage density per unit area. As a result some of the streams which were earlier of second order have turned up to be first order streams. Larger streams are prone to very less changes in their course of flow, direction of flow etc.

2. LAND USE / LAND COVER MAP

   Fig.5.1, Land use / Land cover Map of 2016

   Fig.5.2, Land use- Land cover Map of 2018

   Land Use involves the activities of Human Beings that are directly related to the land. It is the management and modification of the natural surrounding or environment into man-made environment such as settlements etc. Land Cover is basically all the materials on the surface of the earth. It describes the vegetation, water resources, bare ground etc. on the surface of the land. LU/LC maps are most important in river basin planning, river quality assessments, preparation of recreational impact statements, on-shore and off-shore impact development, urbanization studies etc. (U.S. Department of the Interior/Geological Survey

   1982). The LU/LC map can be prepared from the remote sensed data from aircrafts and satellites. These can also be prepared from the field surveys carried out. The Fig.8 shows the LU/LC map of the study area which depicts protected forest, unclassified forest, cultivated area, grass lands, waste lands, built-up areas and water bodies. A comparison of the two land use land cover maps of two different periods reveals an increase in the built up land depicting the growth of urbanization. The two towns at the end of SH 37 i.e. Madikeri and Somvarpet are emerging as centers of growth and development. Significant development is noticed in the adjoining areas as well. Commercial buildings such as homestays constructed on top of unstable slopes have been dragged downstream along with the soil / suffered severe damage at the time of flashfloods and landslides. This points towards the need to adopt sustainable building practices and norms. Natural drainage pathways have also been blocked at many places as a result of excess construction.

3. CONTOUR MAP

   Fig.6, Contour Map of study area

   A contour is a line drawn joining the points of equal elevation. If a contour line represents 100 m then it implies all the points in that line are of the same 100 m elevation from the MSL. A contour map is a topographic map on which the shape of the land surface is shown by contour lines, the relative spacing of the lines indicating the relative slope of the surface. In our study the contour map is extracted from the DEM data in ArcMap. Here the contour value varying every 100m interval is extracted.

4. TIN MAP

   Fig.7, TIN Map of study area

   TIN stands for triangulated irregular network. A TIN is a digital data structure used in a geographic information system (GIS) for the representation of a surface. A TIN is a vector-based representation of the physical land surface or sea bottom, made up of irregularly distributed nodes and lines with three-dimensional coordinates that are arranged in a network of non-overlapping triangles. TINs are easily derived from the elevation data of a rasterized digital elevation model (DEM) in ArcMap.

5. SLOPE MAP

   Fig.8, Slope Map of study area

   A map indicating te topography of an area along with an analysis of topographic features as they have influenced and may continue to influence land development. Slope can be defined as the angle, inclination, steepness, or gradient of a straight line. Slope often is used to describe the steepness of the ground’s surface. The slope of the area is influenced by number of environmental conditions that in turn affect the drainage basin. A steep slope may increase the runoff whereas the gentle or even slope may reduce runoff drastically and also enhance the infiltration in the presence of porous rocks. The slope varies from 0º-52º in our area of study.

   From the above slope map it is understood that our area of interest has relatively gentle slopes. But these slopes along with the loose top soil and heavy rainfall paved the way to landslides at frequent intervals along the state highway. The 45 and more degrees sloppy hillock terrain is not conducive for any construction activity as the fine grade nature of the surface soil is slippery in nature and with heavy rains the possibilities of the upper layer being washed off are more opined by Prof. H. Gangadhar Bhat in his interview with Hindustan Samachar on 14th September 2018.

   The area of interest is classified into landslide hazard zones based on the slope map as follows:

   31.59 51.63	Very high hazard zone
   25.92 31.59	High hazard zone
   21.87 25.92	Moderate hazard zone
   18.42 21.87	Low hazard zone
   0 18.42	Very low hazard zone

6. ASPECT MAP

   Fig.9, Aspect Map of study area

   Aspect values indicate the directions of the physical slope face. The compass direction that a topographic slope faces which is usually measured in degrees from north is the slope aspect. It is the orientation of a landform slope relative to the cardinal points such as southern aspect, northern aspect etc.

7. HILL-SHADE MAP

   Fig.10, Hill shade Map of study area

   The hill-shade function obtains the hypothetical illumination of a surface by determining illumination values for each cell in a raster. It does this by setting a position for a hypothetical light source and calculating the illumination values of each cell in relation to neighboring cells. It is often useful to use a hill-shade raster to show terrain to support other information in a map such as a thematic overlay like soils. It increases the perception of depth in a 3D surface and for analysis of the amount of solar radiation available at a location.

8. GEOMORPHOLOGICAL MAP

   Fig.11, Geomorphological Map of study area (with limited field check)

   In order to study about the geomorphology of the area we have classified the area of interest into different classes such as denudation landforms, alluvial plane, point bar, hill slope etc.

   Geomorphological maps characterize the relief of the earths surface according to its physiognomic features (its morphometric and morphography) and according to its origin and age. Geomorphological mapping is a preliminary tool for land management and geomorphological risk management, also providing baseline data for other sectors of environmental research such as landscape ecology, forestry or soil science (et.al, Otto and Smith, 2013). Our study area is dominated by alluvial plane and denudation landforms. In geology denudation involves the process of wearing away of the earths surface by moving water, by ice, by wind, and by waves, leading to a reduction in elevation and in relief landforms and of landscapes. The presence of denudation landforms owe to the persistence of intense weathering activity in the area and alluvial plane account for the fertility of the soil and its suitability for agriculture. A point bar is a depositional feature made of alluvium that accumulates on the inside bend of streams and rivers below the slip-off slope. Point bars are found in abundance in mature or meandering streams. In our area of interest point bars are not abundant but are indeed present at meandering points. And at few points along the river course channel bars have also been noticed. Sedimentation rates on point bars and on the flood plain indicate two relatively distinct stages of floodplain alluviation et.al, Nanson, 1980).

9. NORMALIZED DIFFERENCE VEGETATION INDEX (NDVI)

   Fig.12.1, NDVI Map of study area of 2016

   Fig.12.2, NDVI Map of study area of 2018

   Comparing the NDVI Maps of 2016 and 2018, the highest values of NDVI increases from 0.546 to 0.594 showing the presence of relatively healthy vegetation. NDVI for vegetation ranges from 0.3 to 0.8, with the larger values representing greener surfaces. Bare soils range from about 0.2 0.3. Thus NDVI provides an estimate of vegetation health and a means of monitoring changes in vegetation over time.

10. NORMALIZED DIFFERENCE BUILT-UP INDEX (NDBI)

    Fig.13.1, NDBI Map of study area of 2016

    Fig.13.2, NDBI Map of study area of 2018

    The NDBI Value is noticed to increase from 0.183 to 0.227 from 2016 to 2018, which shows an increase in urban development. NDBI method is a worthwhile method for mapping urban land. NDBI enables built-up areas to be mapped at a higher degree of accuracy and objectivity. However it is not possible to distinguish industrial, commercial and residential areas in the process (et.al, Zha, Gao,Ni, 2003).

11. GENERAL CAUSE FACTORS:

    Here are some general causes which acts as triggering factors of the landslides: [Source of reference – A Note on the Preliminary Post Disaster Investigation of Landslides Occurred around Madikeri, Kodagu District, Karnataka]

    1. Rainfall: Most of landslide and subsidence were initiated between 15th -17th August 2018. During this period, from 10th August -17th August (8 days), Madikeri Taluk experienced severe rainfall as high as 800 -850 mm, which is almost 25% of the total average annual rainfall (approx. 3400 mm) for last 20 years. This high amount of rainfall acted as triggering factor for the landslides and subsidence as the slope forming material became over saturated by water resulting in increase of pore water pressure and decrease of internal friction.
    2. Modification of Natural Slopes: Modification of the natural slopes is one of the major causal factors for slope failures. In the study area, high angle (vertical or near vertical) slope cut for road construction; rapid slope modification for construction of infrastructure like houses, hotels, home stays etc.; large scale slope modification for plantation (especially coffee) are very common. These slope modifications actually decrease the slope stability and later with the high rain water percolation the slope becomes unstable and prone to failure.
    3. Blockage of Natural Drainage: In the field area, at many places, due to modification of slopes, natural drainage has been blocked. But during high water precipitation in monsoon, huge volume of water flowed from upslope towards downslope but in many places the courses were blocked due to either construction of houses or plantations.
    4. Clean water flow in force from inside the slope: In the field area, at many places, it was observed that considerable volume of very clean water was coming out in force from the middle parts of the landslides. This continuous flow of the water in force inside the slope will decrease the stability of the slope and eventually will act as a triggering factor for slope failure.
    5. Water Tanks / Ponds in Coffee Estates: Another very important reason is the creation of water tanks / ponds throughout the slope in the plantation estates for watering the plants in the lean season. During the heavy rainfall, all these tanks / ponds got overflowed. There are also high possibilities of seepage from within the tanks directly within the slope. This wate along with the rain water will increase the pore water pressure of the slope and decrease the internal friction of the slope forming materials which ultimately lead to the slope failures.
    6. Flash flood due to temporary dam formation in nala: In the field area, it was observed at many places that the nala courses became very wide at places and is filled with large tree trunks, boulders along with high amount of debris. These tree trunks and boulders along the accumulated debris blocked the course of the nala forming a temporary dam structure. These dam structures might have blocked the water from upstream for several hours. As the water level kept on increasing, the water pressures became so severe that the temporary dam breached and a high volume of

       water along with debris and tree trunks released at force to form a flash flood.

    7. Geological Causes: The main rock type found in the study area is Granite gneiss and garnet silimanite graphite schist which are prone to weathering. The area of study is prone to high amount of chemical weathering. The rate and amount of top soil formation is more and also the soil being characterized as very fine grained alluvial soil. In three days of heavy rainfall the top soil got heavily saturated thus increase in self weight of the soil mass. Also the layer beneath the top soil is clay layer which again has less resistance to sliding.

IV. CONCLUSION

Landslides in hilly regions are complicated in nature. The problem of landslides increases year by year and is strongly influenced by developmental work, particularly road widening, tourism development and many new infrastructural projects. After our study we have come to the conclusion that the main reason for the sudden calamity has been excessive rainfall of over 800mm received by the district in a few number of days. Also the ecology of the area has been severely disturbed by construction activities, construction of water tanks on top of steep slopes, construction of homestays and blockage of natural drainage pathways leading to increased infiltration of the blocked water into the soil strata thereby again contributing to an increase of saturation of top soil. Intrusive rocks such as dolerite dykes and metamorphic rocks such as peninsular gneiss constitute the lithology of the region. These rocks normally being classified as hard rocks are found to have undergone extensive chemical weathering which in turn also accounts for the highly fertile top soil and large-scale plantation agriculture. Thus to recapitulate excessive rainfall, highly fractured rock, fine grained soil, anthropogenic activities such as excess deforestation are the major causative parameters of the calamity.

From the point of view of our findings and results the disaster that happened cannot be classified completely as a manmade disaster, rather it has been caused due to a combination of both geohazard and anthropogenic interventions. We prefer to call it as a geohazard triggered by anthropogenic causes. Some of the permanent restoration works that can be implemented in Kodagu include hydro seeding, soil nailing, construction of retaining structures etc. However none of the restoration method can be overemphasized over the restoration of the natural ecology.

The disaster is indeed a wakeup call for the government as well as the residents of the area to conserve the ecology and to prevent further exploitation of the natural resources. As of now the environment of the hilly district is in a very critical condition. If suitable permanent restoration measures as well as conservatory norms are not implemented, we can definitely witness the recurrence of a similar disaster of a greater intensity in the upcoming monsoon seasons as well. The catastrophe that occurred is a clear justification of the Gadgill committee recommendations that was earlier neglected or corrupted

by the government. According to Gadgill committee report the entire Western Ghats were grouped into three different levels of ecological sensitive zones wherein most of the developmental activities were restricted to different extents. The implementation of these recommendations could sustain the deteriorating ecology of the Ghats by saving the natural resources from their over exploitation. The negligence of the government towards taking up these guidelines paved way to further degradation of the Western Ghats ecology creating suitable circumstances for such natural calamities to occur.

REFERENCES

1. YZHA, J.GAO, S.NI (2001) Use of Normalized Difference Building Index in automatically mapping urban areas from TM imagery. International jouranal for Remote sensing, 2003, Vol-24, No.3, 583- 594.
2. Bora Mukul Chandra (2010) The flood situation of Assam A case study. BALWOIS 2010- Orchid, Republic of Macedonia 25, 29 May 2010.
3. Haigh Martin and J.S Rawat (2012) Landslide Disaster: seeking causes: A case study from Uttarakhand , India, Chapter 18, pp 218 253, in krecek J , Haigh M , Hofer T and Kurbin E (eds).Management of mountain watersheds.
4. D.N Vinutha, K Shridhara Raje and M.R Janardhana (2014) Landform studies and geomorphological mapping of a part of coorg district, Karnataka state. International journal of Geology, earth and Environmental science ISSN: 2277-2081.
5. E Palanivel , J Saravanavel , S Gunasekaran (2015) Disasters management. Chapter-III: Classification of Disasters. Chapter-IV: Disaster Vulnerable Area Mapping, Mitigation, Damage Assessment and Management-Application of Geomatics Technology.
6. Jan christoph Otto and M.K.J Smith (2013) Geomorphological Mapping – British Society for Geomorphology Geomorphological Techniques, Chap.2, Sec.6.
7. Bhatt Vijay Dev (2013) Spatial Mapping Of Landslides By Using Arcgis And Google Earth: A Case Study Of Chure Region In Kanchanpur District, Far-Western Nepal, Tribhuvan University.
8. M. Pierson Suzanne, J. Rosenbaum Barbara, D. McKay Lucinda, and G. Dewald Thomas (2008) Strahler Stream Order and Strahler Calculator Values in NHDPlus, SOSC Technical paper.
9. 12. Megokedono Vakha, M Nagaraju , K.G Asha Manjari (2016) Land use land cover mapping using Geo-informatics of Kodagu District, Karnataka, International Journal Of Geomatics And Geosciences Volume 7, No 1, 2016.
10. R. Anderson James, E. Hardy Ernest, T. Roach John and E. Witmer Richard (2001) A Land Use And Land Cover Classification System For Use With Remote Sensor Data – Geological Survey Professional Paper 964.
11. P.B Durgin, Landslides and the weathering of granitic rocks (1977) Geological Society of America Reviews in Engineering Geology, Vol.3.pp.127 132, 1977.
12. Sibomana Alexis and H. Gangadhara Bhat (2018) Drainage network extraction from ASTER G DEM 1 arc second resolution using D8- model for Kagunzi watershed in Kibira National Park North- Western part of Burundi