*"Report of the Western Ghats Ecology Expert Panel - Part I - 17. Mining in Goa : 17.4 Recommendations : Appendices : Appendix 1: Methodology employed in generating and interpreting the Western Ghats Database and assigning ESZs : The following data sets were used for geo spatial analyses.

Opinion
       11/10/2018
                 1458.

Sub :-

*"Report of the Western Ghats Ecology Expert Panel - Part I - 17. Mining in Goa : 17.4 Recommendations : Appendices  : Appendix 1: Methodology employed in generating and interpreting the Western Ghats Database and assigning ESZs : The following data sets were used for geo spatial analyses.

Ref :

17.4 Recommendations :-

Appendices :-

Appendix 1: Methodology employed in generating and interpreting the Western Ghats Database and assigning ESZs
The following datasets were used for geospatial analyses.

I. Data Sets : -

1. Western Ghats boundary (shape file) obtained from Dr. Ganeshaiah, Member, WGEEP
2. India states, districts, talukas (shape file ) source : DIVA-GIS (http://www.diva-gis.org/)
3. Shuttle Radar Topographic Mission (SRTM) data for India (TIFF) at 90 m resolution.
4. Data on endemic plants, IUCN Red list Mammals, percent forest, unique evergreen elements, forest with low edge: (from Das et al., 2006) 25k grid (shape file)
5. Forest types of India (TIFF)
6. Protected Areas of Western Ghats Cover (shape file) Source: FERAL
7. Elephant Corridors of Western Ghats Cover (shape file) Source: Prof R Sukumar, CES, and WTI.
8. Endemic vertebrate data of Western Ghats Cover (Spread sheet) Source: Ranjit Daniels
9. Endemic Odonata data of Western Ghats Cover (shape file) Source: ZSI
10. Enhanced vegetation index of MODIS for North Maharashtra and Gujarat
11. Riparian Forests derived through drainage and forest cover
12. Important Bird Areas (IBAs) as point coverages

Of these, data sets 1–5 and 8–12 were used for the geospatial analyses. For North Maharashtra and Gujarat, Enhanced Vegetation Index (EVI) of MODIS was used as the forest vegetation data were not readily available.

Use of Free and Open Source Software : -

Free and Open source geospatial tools (www.osgeo.org) were extensively used as given below :-

Desktop GIS: Open jump, QGIS, SAGA, DIVA-GIS
Database: PostgreSQL/ PostGIS
Web GIS: OpenGeo Suite which is a complete web platform based upon Open Geospatial Standards (OGC) which includes GeoServer (GIS Server), PostgreSQL/PostGIS(Database), Geo Web Cache (Cache Engine), Geoexplorer (for Visualization of WMS layers), GeoEditor (Online editing geospatial data), and Styler (Online styling of the data).
A web enabled searchable database has been a major contribution of this short-term project. In addition, through UNICODE, local language adoption has been showcased using Marathi as an example.
--
In addition, using methods of spatial analyses on large landscape level data, an attempt was made to arrive at the relative importance of these seven attributes. This has been done using a programme called Spatial analyses in Macro Ecology (SAM) . However, this has been done only on a preliminary exploratory basis to showcase one possible way of reducing the dimensionality of the factors involved. Not much headway was made with this approach due to several operational constraints.

II.  Data Cleaning Process : -

a. 5 minute x 5 minute grid file generation for Western Ghats Cover (shape file) using Vector Grid plugin of QGIS
b. 1 minute x 1 minute grid file generation for Western Ghats Cover of Goa state (shape file) using Vector Grid plugin of QGIS
c. Rasterization of each attribute of ATREE data by applying Surface method using Rasterize (Vector to Raster) plugin of QGIS
d. Generated slope map in TIFF format using GDAL library
e. Generated shape files for following classes in Endemic Vertebrate data (Ranjit Daniels, 2011)
 Amphibians
 Birds
 Reptiles
 Fish
 Endemic Odonata (ZSI, 2011)

III. Uploading datasets into database : -

All the available and generated datasets were uploaded to the PostgreSQL/PostGIS database using QGIS as below. The vector datasets were uploaded to the database using the SPIT plugin of QGIS while raster datasets were uploaded using Load Raster to PostGIS plugin of QGIS. In case of Raster dataset, the data was stored into 64 x 64 blocks.

V.. Vector/Raster analysis using PG Raster of PostGIS :-

a. Vector/Raster analysis was done for elevation values from SRTM data using WKT Raster Queries. Following is the sample query for it.
Sample Query : -
Create table <table name> as SELECT e.id,test.val, ST_Intersection(test.geom, e.geometry) AS gv FROM (SELECT (ST_DumpAsPolygons(ST_SetBandNodataValue(rast, 0))).geom, (ST_DumpAsPolygons(ST_SetBandNodataValue(rast, 0))).val FROM <Raster_table_name>) as test, <Grid_table_name> as e WHERE ST_Intersects(test.geom, e.geometry);

VI. Grouping and averaging of pixel values based upon grids :-

Thereafter, average elevation values were calculated for each 5' x 5' grid for each state in the Western Ghats and considered as a parameter.
The steps 4–5 were performed for parameters such as maximum slope values, endemic plants, iucn max, unique percent, comp3 percent, forest percent values, area of riparian forest (see explanation of parameter below) for each 5' x 5' grid for each state in the Western Ghats Cover.

VII. Ranking the parameters generated :-

Assigned ranks for the following 8 parameters

a. Endemic plants : Number of endemic plant species
b. IUCN_max: Number of IUCN Red listed mammal species
c. Unique percent: Percentage of area covered by unique evergreen ecosystems
d. Comp3 percent : Percentage of area covered by relatively undisturbed forest with low edge
e. Forest percent: Percentage of forest area
f. Elevation
g. Slope
h. Riparian Forests/Vegetation

As there is an ecological gradient from north to south in the Western Ghats with changes in diversity and species richness as well as physical features, a normalization for every state was done for these parameters. Thus, scores were normalized for each state. For instance, the highest recorded altitude in a given grid in a state was given the maximal score and all other grids in that state were ranked in relative fashion. After normalization ranks were assigned on a scale from 1 to 10 based on the maximum value of each parameter for each state.

VIII.  Average of the ranks for all parameters :-

Subsequent to the rank generation, the average of the ranks for all parameters were calculated. If, for a grid, there is data for only for 5 parameters out of 8 parameters, then dividing the sum by the number of parameters assessed took care of the problem of data available for variable numbers of parameters per grid.

IX. ESZ assignment algorithm :-
1. We treat Western Ghats regions of each state separately
a. Existing Protected Areas are treated as a fourth separate category
b. ESZ1, ESZ2 and ESZ3 status are assigned only to grids outside existing Protected Areas
c. ESZ1 status are assigned only to such grids as have a score at least equalling, or higher than the lowest scoring grids falling within existing Protected Areas
d. The extent of existing Protected Areas plus ESZ1will not normally exceed 60% of the total area
e. The extent of ESZ3 will normally be around 25% of the total area

With these stipulations, we adopt the following procedure : -
Let p be the percentage of area falling under existing Protected Areas
Let x be the percentage of area assigned to ESZ1
Let y be the percentage of area assigned to ESZ2
Let z be the percentage of area assigned to ESZ3
Obviously, p+x+y+z = 100

Now, we can visualize three scenarios in terms of value of p; [1] p>75, [2] 60<p<75, and [3] p<60. Normally p<60 will hold, but logically we must allow for the first two as well.
[1] p>75: In this case, all areas outside existing Protected Areas will be assigned to ESZ3. No grids will be assigned to ESZ1 or ESZ2, as existing Protected Areas themselves exceed 75% of the region. x=0, y=0, z= (100–p);
so that x+y+z+p= 0+0+(100–p)+p=100

[2] 60<p<75: In this case, we will assign the lowest scoring 25% of grids to ESZ3 and the balance grids to ESZ2. No grids will be assigned to ESZ1, as existing Protected Areas themselves exceed 60% of the region. Then, x=0, y=(75–p), z=25 leading to
x+y+z+p= 0+(75–p)+25+p=100

[3a] p<60: This will be the normal case. In this case, we will assign the lowest scoring 25% of grids to ESZ3. The balance of (75–p) has to be assigned to ESZ1 and ESZ2 such that p+ESZ1=60. Since we accept that existing Protected Areas and ESZ1 should not exceed 60%, we have to assign all of the top scoring 60% grids that are outside existing Protected Areas to ESZ1, provided that the lowest score amongst these at least equals or is higher than the lowest score of the grids falling within existing Protected Areas.
So, in this scenario of 60<p<75; x=(60–p), y=15, z=25, and
x+y+z+p= (60–p)+15+25+p=100.

[3b] One more special case, has to be considered for this scenario of p<60, namely that equating the lowest score of the grids falling within existing Protected Areas to the lowest score of the grids assigned to ESZ1 does not assign enough grids to ESZ1, so that (p+x)<60. In that case, the balance of the top scoring 75% grids that are outside existing Protected Areas, and grids assigned to ESZ1, will be assigned to ESZ2. So, y=75–(p+x), and will be more than 15%.
Again, x+y+z+p= x+75–(p+x)+25+p=100

[4] An additional, score assignment device has been introduced. When we want to select some specific percentage of grids, say, lowest 25%, setting the threshold to a specific integral score may not yield the desired result. Then, we rank the parameters used to generate the scores in the order of their importance, and rework the scores by ignoring the least important parameters till roughly the desired percentage, say between 22 to 28, is reached.

To make administration easy, the ESZ are extrapolated and reported for talukas. The assigned ESZ level to the taluka is the ESZ that covers the largest fraction of the taluka.
In the case of Goa, because of its size and the use of 1 minute x 1 minute grids, ESZs are not reported for whole talukas, but by grids within talukas.

The method is illustrated for Goa : -

a. A WG database for Goa is prepared as discussed above
b. The parameters are ranked on a 1-10 scale, with lowest at 1 and highest ecological significance at
c. Composite scores – average for each grid- are calculated
d. For arriving at ESZs, the grid scores were treated thus: All grids having PAs are excluded for arriving at the ESZ1. Since these grids also have scores, a guiding strategy for demarcation of ESZ1 is the range of scores for PAs of a given state. Thus the average minimum threshold for Goa PAs is 4.92. Hence all grids having a score of above 4.92 get assigned to ESZ1.Thus 11 grids out of a total of 55 grids make the cut (20%). The grids with PAs are 21 in number and account for 38% of the total grids. ESZ1 and PAs together constitute 58%. the lowest quartile (approx. 25%) of these scores for grids was computed. For Goa , this score is 3.14 which means all grids below this core are assigned to ESZ
3. For Goa there are 12 grids under ESZ3 , which constitute about 22% of the area. The balance of grids are assigned to ESZ2. These are 11 in number (20%, a deviation of 5% from the suggested 15% of area).

X. Outputs :-

The results obtained are presented as

a. A spatial depiction of ESZs grid-wise as well as taluka-wise and displayed on a colour palette , with Green showing ESZ1, Red showing ESZ2 and yellow showing ESZ3.
b. Percent grids for a given score for each state both in a tabular and graphical notation
c. Riparian forest scores for each state and in different elevation zones
d. 1' x 1' grid analysis for Goa to incorporate the results of the Goa Regional plan
e. A Web GIS application

XI. Information and Data Sources :

a. Habitat related information in the form of shape files for parts of Mahrashtra, Karnataka, Kerala and Tamil Nadu: Mr Kiran , Arundhati Das, V Srinivasan and Dr Jagdish Krishnaswamy of ATREE Additional data from Ravindra Bhalla of FERAL and Bhaskar Acharya of CEPF
b. Dr RJR Daniels of Care Earth: point locations of mammals, reptiles, birds, amphibians and fishes
c. Dr K A Subramanian , ZSI: point locations of Odonata
d. Prof R Sukumar: information on elephant corridors
e. Dr K N Ganeshiah: Western Ghats boundary
f. Dr P S Roy, Director, Indian Institute of Remote sensing, Dehra Dun: habitat information and shape files for Gujarat and Maharashtra
g. Dr Bharucha and Shamita from BVIEER, Pune: data on parts of Maharashtra
h. Dr K S Rajan , Open Source Geospatial Foundation – India chapter and IIIT, Hyderabad : geospatial statistical analyses
i. Dr P V K Nair, KFRI: assistance in analyses for Kerala
j. Santosh Gaikwad, Siva Krishna, Ravi Kumar, Ch.Appalachari, Sai Prasad of SACON: GIS work.

NEXT : Appendix 2: Proposed assignment of various Western Ghats Talukas to ESZ1, ESZ2 and ESZ3

To be continued ..

OPINION :

1.WHILE WRITING THIS POST AFTER WIDE READING OF THIS REPORT, THRILLED BY THE DEDICATION ON THE SUBJECT BY SRI MADHAV GADGIL;

2.THE REPORT WAS SUBMITTED ON 2011;

3. EXCEPT KERALA ALL OTHER STAKE HOLDERS OF WESTERN GHATS, STARTED IMPLEMENTING THE RECOMMENDATIONS;

4.KERALA IS PLAYING PARTY+VOTE BANK+ COMMUNAL POLITICS, THE PRESENT LDF GOVT AS WELL AS PREVIOUS UDF GOVT RELUCTANT TO IMPLEMENT;

5.UNDER VARIOUS CASTE, RELIGION, AND OTHER MAFIAS INFLUENCE / PRESSURE; PARTY VOTE BANK POLITICS IS IMPORTANT THAN CONSERVATION OF WESTERN GHATS - TO KERALA POLITICIAN CUM PEOPLE

NO FLOOD CAN TEACH A LESSON TO THESE IDIOTS ..

***
JAIHIND
VANDEMATHARAM