252
administrative boundary layer of the study area, the image was given the desired
shape. The VCI images were classified in three classes: severe, moderate, and normal drought condition in Arc GIS 10.1 software on the basis of VCI values. VCI
values of each districts in different years were derived from VCI image through
means of spatial average statistics or zonal statistics, the values obtained for each
district for each year were then tabulated in MS excel, and VCI trend were drawn as
a line graph.
There are two rainfall stations in entire study area, and the rainfall data is not
regular as well as costly, so CHIRPS data (Climate Hazard Group Infrared
Precipitation with station) was chosen for computing RAI. CHIRPS is a product
derived through collaboration among scientists at the US Geological Survey (USGS)
Earth Resources Observation and Science (EROS) Center. It is reliable, up to date,
and more complete datasets which have already been used for a number of early
warning studies such as seasonal drought monitoring (Rojas 2018). CHIRPS data is
a 30+ year quasi-global rainfall dataset. Spanning 50°S-50°N and at all longitudes,
this data is available from 1981 to near present. CHIRPS incorporates 0.05° resolution satellite imagery with in situ station data to create gridded rainfall time series
for seasonal drought monitoring. As of February 12, 2015, version 2.0 of CHIRPS
is complete and available for estimating rainfall variation in space and time (Funk
et al. 2015). CHIRPS’s rainfall maps are useful, especially for places where surface
data is sparse (Funk et al. 2015). However, estimates derived from satellite data
provide areal average that suffers from biases due to complex terrain and often
underestimate the intensity of extreme precipitation events (Funk et al. 2015).
In line with the methodology followed by Kundu et al. (2016), this study also
uses CHIRPS rainfall data product for Bankura, Purulia, Purba, and Paschim
Medinipur district which were downloaded from ftp://ftp.chg.ucsb.edu/pub/org/
chg/product/chirps-2.0/global daily/tifp05 for over 20 years (1998–2017) for July
and August, to enable assessment of the spatiotemporal dynamics of meteorological
drought and its relation with agricultural drought, that is RAI and VCI respectively.
This data was then processed through ENVI Classic 4.7 software. At first the
monthly dataset was generated by layer stacking the daily downloaded datasets.
This stacked layer was masked using the ROI (region of interest) masking layer to
obtain the shape of the study area. Average of monthly rainfall was successively
derived using formula of computing average or mean in band math option [Formula:
(day1 + day2…. +day31)/31, where month is of 31 days]. The monthly dataset of
July and August was again averaged for 20 years. Then the mean and standard deviation of the 20 years average were computed. After post-processing of rainfall datasets, RAI values were derived from the images through Arc GIS10.1 using the
formula given in Eq. (14.2). Zonal statistics in Arctools were used to generate district level RAI data for each year. The derived values were tabulated in excel, and
line graph was drawn to represent the RAI trend for the four districts from 1998 to
2017. On the basis of Rooy 1965 the rank of precipitation values were calculated to
be positive and negative anomalies, and accordingly RAI values were interpreted.
Equation 14.2 represents computation of deviation in rainfall from long-term average (Dutta et al. 2013, 2015; Padhee et al. 2017).
M. Dey et al.
administrative boundary layer of the study area, the image was given the desired
shape. The VCI images were classified in three classes: severe, moderate, and normal drought condition in Arc GIS 10.1 software on the basis of VCI values. VCI
values of each districts in different years were derived from VCI image through
means of spatial average statistics or zonal statistics, the values obtained for each
district for each year were then tabulated in MS excel, and VCI trend were drawn as
a line graph.
There are two rainfall stations in entire study area, and the rainfall data is not
regular as well as costly, so CHIRPS data (Climate Hazard Group Infrared
Precipitation with station) was chosen for computing RAI. CHIRPS is a product
derived through collaboration among scientists at the US Geological Survey (USGS)
Earth Resources Observation and Science (EROS) Center. It is reliable, up to date,
and more complete datasets which have already been used for a number of early
warning studies such as seasonal drought monitoring (Rojas 2018). CHIRPS data is
a 30+ year quasi-global rainfall dataset. Spanning 50°S-50°N and at all longitudes,
this data is available from 1981 to near present. CHIRPS incorporates 0.05° resolution satellite imagery with in situ station data to create gridded rainfall time series
for seasonal drought monitoring. As of February 12, 2015, version 2.0 of CHIRPS
is complete and available for estimating rainfall variation in space and time (Funk
et al. 2015). CHIRPS’s rainfall maps are useful, especially for places where surface
data is sparse (Funk et al. 2015). However, estimates derived from satellite data
provide areal average that suffers from biases due to complex terrain and often
underestimate the intensity of extreme precipitation events (Funk et al. 2015).
In line with the methodology followed by Kundu et al. (2016), this study also
uses CHIRPS rainfall data product for Bankura, Purulia, Purba, and Paschim
Medinipur district which were downloaded from ftp://ftp.chg.ucsb.edu/pub/org/
chg/product/chirps-2.0/global daily/tifp05 for over 20 years (1998–2017) for July
and August, to enable assessment of the spatiotemporal dynamics of meteorological
drought and its relation with agricultural drought, that is RAI and VCI respectively.
This data was then processed through ENVI Classic 4.7 software. At first the
monthly dataset was generated by layer stacking the daily downloaded datasets.
This stacked layer was masked using the ROI (region of interest) masking layer to
obtain the shape of the study area. Average of monthly rainfall was successively
derived using formula of computing average or mean in band math option [Formula:
(day1 + day2…. +day31)/31, where month is of 31 days]. The monthly dataset of
July and August was again averaged for 20 years. Then the mean and standard deviation of the 20 years average were computed. After post-processing of rainfall datasets, RAI values were derived from the images through Arc GIS10.1 using the
formula given in Eq. (14.2). Zonal statistics in Arctools were used to generate district level RAI data for each year. The derived values were tabulated in excel, and
line graph was drawn to represent the RAI trend for the four districts from 1998 to
2017. On the basis of Rooy 1965 the rank of precipitation values were calculated to
be positive and negative anomalies, and accordingly RAI values were interpreted.
Equation 14.2 represents computation of deviation in rainfall from long-term average (Dutta et al. 2013, 2015; Padhee et al. 2017).
M. Dey et al.