Presentation by Chandrashekhar Biradar and team.
16-18 October 2019. Hyderabad, India. TRUST: Humans, Machines & Ecosystems. This year’s Convention was hosted by The International Crops Research Institute for the Semi-Arid Tropics (ICRISAT). The Platform is led by the International Center for Tropical Agriculture (CIAT) and the International Food Policy Research Institute (IFPRI).
Mapping suitable niche for cactus and legumes in diversified farming in drylands
1. Mapping suitable
niche for cactus and
legumes in diversified
farming in drylands
Chandrashekhar Biradar
and team
2. #BDPHYDERABAD2019
Current Diets vs Planetary Health
EAT Lancet Report
Moving from narrow sense economic benefit to a new
ecologically sound functional system for well being…
3. #BDPHYDERABAD2019
EAT Lancet Report
…with diversified cropping systems, conservation,
rotation, nutrition focus >> “more health per acre”
Dryland
Cereals
Dryland
Pulses
Dryland
Livestock
Current Diets vs Planetary Health
5. #BDPHYDERABAD2019
Daal/Falafal
Water used 1,250 liters
Chicken
4,325
Mutton
5,520
Beef
13,000
Changing diet pattern >> cropping systems
Crop diversification for future smart foods
There is a need for paradigm
shift from more calories per
acre to more health per acre.
>> Sustainable living
6. #BDPHYDERABAD2019
6
Crops of resilient systems
• Diversified farming systems
• Sustainable food and nutrition
• High resource use efficiency
• Rebuilding living soils
• Crops of resilient systems
Crop type Total Soil loses
Bare soil 29.10
Cotton 10.91
Cereals 5.94
Ceraisl + beans 3.93
Opuntia ficus-indica 1.98
Perennial grass 0.03
7. #BDPHYDERABAD2019
Nutrition Rich Food: Fruits, Juices, Vegetables, Pickles, Fodder
Why cactus is important?
Cosmetic oilSoaps/shampoo Natural die
anti-wrinkle,
and anti-age purposes
8. #BDPHYDERABAD2019
Cactus with date palm
Cactus with cluster bean Degraded wastelands
Cactus with citrus Cactus with chickpea Cactus/trees/crop
Cactus in different farming systems
Cactus with Napier Cactus with Barley
9. #BDPHYDERABAD2019
Image
download
Atmospheric
correction
Save RED and
NIR band to
tmp folder,
names of files:
*DOY_RED.tif
*DOY_NIR.tif
Read two
tif files
from tmp
folder, do
NDVI
calculation
Save NDVI file to folder
“basename_DOY_ndvi.tif”
Add column to
existing csv
Module-1
Module-3
Module-2
Shapefile
all_fields_fergana
.shp
Field
:
“ID”
Do 10-day
interpolation,
add columns
GEOTIFF of field
IDs
Zonal
statistics
Pre-
existing
data
Empty CSV with
field IDs in folder
named after
recent year?
Do
classification
Update columns
in CSV
(classification,
probabilities)
Output:
Saveto“annual”folders
Processing
Inputorexisting
database
Save final csv for
classification
“fergana_croptype_YY
YY.csv”
Save final csv for
interpolated NDVI
“fergana_NDVI_ts_int
erpolated_YYYY.csv”
Save updated,
historic data
base for VCI as
CSV or R Object?
“historic_vci_db_YYY
Y.csv”
Do table join: Update
annual shapefile with
classifications,
probabilities, and VCI
indicator (seaosnal and
end-or-seaosn each)
1x1 km grid
as shapefile
Compute MIN
and MAX NDVI
multi-annual
reference for
each grid
Update columns (VCI
deviation +/-)
Save final csv for VCI
“fergana_vci_YYYY.csv
”
Front-end
system
File Geodatabase
Front-end
Web-mapping
service (hosted
or referenced)
Front
end
Pooled data sets
and RF models
(seasonal and
annual)
Mean
Save atmospherically
corrected images file to
folder
“basename_DOY.tif”
Big-data, Machine Learning and AI
SVM, BT, LR, RF, DT, MLP
Multi-mode classification algorithms
10. #BDPHYDERABAD2019
#/km2
Dynamics of Cropping Systems
▪ Integrated Agro-Ecosystems
▪ Sustainable Intensification and Diversification
▪ Input Use Efficiency-Conservation Agriculture
▪ Thematic Land-Water-Climate Resilience
Agricultural
Intensification
Cropping Intensity
Increase in Arable
Land
72%
21%
7%
Length of the crop fallows, start-date, end-date
(Biradar et al., 2015)
13. #BDPHYDERABAD2019
Legumes and Cactus as
integrated crops
From 2000 to current (real-time mapping)
Mapping Realtime
Rice-Fallows
Soil Moisture and
Water Harvesting
Variety Suitability
Agro-Tagging
14. #BDPHYDERABAD2019
Mapping
suitable niche
Rainfall Anomaly
Mean Rainfall Rainfall trend
Suitability based on rainfall variation
• 32% of India is in the ‘high (3%) to moderate suitable (29%)’ category.
• Precipitation anomaly suggests high suitability coerced in western and east-central part.
• A decreasing rate of suitability with increase of aridity was found.
𝑆𝑃𝐼 =
𝑥𝑖 − 𝜇 𝑥
𝜎𝑥
𝑇 𝑚𝑘 =
𝑖=1
𝑛−1
𝑗=𝑖+1
𝑛
𝑠𝑔𝑛 𝑥𝑗 − 𝑥𝑖
𝑆 =
𝑖=1
𝑛
𝑊𝑖 𝑆𝑖
15. #BDPHYDERABAD2019
Precipitation 400-800 mm/yr Optimum
800-1000
mm/yr
Suitable
250-400 mm/yr Less Suitable
< 250 and >
1000 mm/yr
Not Suitable
Min.
temperature
more than 5 °C Optimum
>-5°C to < 5 °C Less suitable
<5°C Not suitable
Max.
Temperature
> 41°C Suitable
5 to 41°C Optimum
< 5°C Less Suitable
Mean
Temperature
> 23°C Less suitable
18-23°C Optimum
15 -18°C Suitable
10 -15°C Less suitable
Less than 10°C Not suitable
Annual
relative
humidity
>60% Optimum
40-60% Suitable
<40% Not suitable
Soil
Salinity
<2 Optimum
2 to 4 Suitable
4 to 7 Less suitable
more than 7 Not suitable
Soil
Texture
clay (heavy/light),
clay loam
Not suitable
Silt/sandy clay loam Less suitable
Loam, sandy loam,
loamy sand
Suitable
sandy Optimum
Soil PH 5 to 8 Optimum
<5 and >8 Not suitable
Soil
organic
matter
1 to 2 Optimal
0.5-1 Suitable
< 0.5 Less suitable
Absent Not suitable
Parameters for Multi-criteria analysis (AHP)
Conventional Approach
Cloud computing: GEE