30
REMOTE SENSINGBASED INFORMATION AND INSURANCE FOR CROPS IN EMERGING ECONOMIES RIICE service Earth Observation component Francesco Holecz – sarmap

RIICE service Earth Observation componentsari.umd.edu/sites/default/files/Franceso Holecz.pdf · • Soil data • Phenological data • Management data Yield estimation ... mobile

  • Upload
    others

  • View
    1

  • Download
    0

Embed Size (px)

Citation preview

Page 1: RIICE service Earth Observation componentsari.umd.edu/sites/default/files/Franceso Holecz.pdf · • Soil data • Phenological data • Management data Yield estimation ... mobile

REMOTE SENSING‐BASED INFORMATION AND INSURANCE FOR CROPS IN EMERGING ECONOMIES

RIICE service

Earth Observation component

Francesco Holecz – sarmap

Page 2: RIICE service Earth Observation componentsari.umd.edu/sites/default/files/Franceso Holecz.pdf · • Soil data • Phenological data • Management data Yield estimation ... mobile

REMOTE SENSING‐BASED INFORMATION AND INSURANCE FOR CROPS IN EMERGING ECONOMIES

Donor, objectives and countries

RIICE is funded by the Swiss Agency for Development and Cooperation.

Key objectives:• Provision of reliable rice production information in major rice

growing areas in South‐East Asia.• To develop a model aiming at improving rice production forecast.• To transfer the appropriate know‐how to national partners.• Setting up sustainable micro‐insurance schemes by developing

insurance solutions covering production shortfalls (e.g. from floodand drought).

Page 3: RIICE service Earth Observation componentsari.umd.edu/sites/default/files/Franceso Holecz.pdf · • Soil data • Phenological data • Management data Yield estimation ... mobile

REMOTE SENSING‐BASED INFORMATION AND INSURANCE FOR CROPS IN EMERGING ECONOMIES

Multi‐stakeholder partnership

International reinsurance market 

provides risk capacity

Local insurance company issues policies and 

administers the insurance claims

Aggregator (rural bank or commune) manages distribution of insurance policies

Government provides insurance premium subsidies (up to 100%) motivated to

a) stabilize farmers incomesb) keep national budget less volatile

Farmer receives insurance coverage against crop loss

SwissRe & AZREprovide insurance solution

IRRI provides yield modellingand know‐how

SDC & GIZ build capacity and facilitate policy dialogues 

sarmap provides remote sensing technologyand know‐how

National partnersreceive appropriate

know‐how 

Page 4: RIICE service Earth Observation componentsari.umd.edu/sites/default/files/Franceso Holecz.pdf · • Soil data • Phenological data • Management data Yield estimation ... mobile

REMOTE SENSING‐BASED INFORMATION AND INSURANCE FOR CROPS IN EMERGING ECONOMIES

The service

ORYZArice growth simulation model

• Meteo data• Soil data• Phenological data• Management data

Yield estimation

• Rice eco‐system map• Crops calender• Administrative units

• Seasonal Area• Start of season date• Leaf Area Index • Seasonal dynamics• Flood damage • Drought damage

MAPscape‐RICESAR & Optical data processing

Earth Observation data

Leaf Area Indexin situ point data

Production 

RIICE answers to three crucial questions:• Where?• When?• How much?

Page 5: RIICE service Earth Observation componentsari.umd.edu/sites/default/files/Franceso Holecz.pdf · • Soil data • Phenological data • Management data Yield estimation ... mobile

REMOTE SENSING‐BASED INFORMATION AND INSURANCE FOR CROPS IN EMERGING ECONOMIES

MODIS 2002 250m 

Landsat‐5 1985‐2012 30mLandsat‐8 2013 30m

Sentinel‐1A/B 2014/2016 20mSentinel‐1C/D 2020 20m

Sentinel‐2A/B 2015/2017 10, 20mSentinel‐2C/2D 2020 10, 20m

Satellite free of charge data are ensured until 2030

Free of charge satellite SAR and optical data

Automated data processing

Page 6: RIICE service Earth Observation componentsari.umd.edu/sites/default/files/Franceso Holecz.pdf · • Soil data • Phenological data • Management data Yield estimation ... mobile

REMOTE SENSING‐BASED INFORMATION AND INSURANCE FOR CROPS IN EMERGING ECONOMIES

Rice imaged by Synthetic Aperture Radar

Ideal rice temporal signature

Temporal signature depends on:

• Eco‐system• Practices• Establishment method• Cycle duration• Biomass & moisture

Page 7: RIICE service Earth Observation componentsari.umd.edu/sites/default/files/Franceso Holecz.pdf · • Soil data • Phenological data • Management data Yield estimation ... mobile

REMOTE SENSING‐BASED INFORMATION AND INSURANCE FOR CROPS IN EMERGING ECONOMIES

© ESRI

All Earth Observation data are transferred, stored, processed and analyzed on the cloud. 

All field data collected by mobile phone, sent to the cloud over mobile or Wi‐Fi network. 

Users access information via a web‐based platform from any internet enabled device. 

Service infrastructure

Page 8: RIICE service Earth Observation componentsari.umd.edu/sites/default/files/Franceso Holecz.pdf · • Soil data • Phenological data • Management data Yield estimation ... mobile

REMOTE SENSING‐BASED INFORMATION AND INSURANCE FOR CROPS IN EMERGING ECONOMIES

Seasonal rice area – Approach X‐band HH time‐series

Page 9: RIICE service Earth Observation componentsari.umd.edu/sites/default/files/Franceso Holecz.pdf · • Soil data • Phenological data • Management data Yield estimation ... mobile

REMOTE SENSING‐BASED INFORMATION AND INSURANCE FOR CROPS IN EMERGING ECONOMIES

Mekong River Delta, 2013 – Where, When, How much

Page 10: RIICE service Earth Observation componentsari.umd.edu/sites/default/files/Franceso Holecz.pdf · • Soil data • Phenological data • Management data Yield estimation ... mobile

REMOTE SENSING‐BASED INFORMATION AND INSURANCE FOR CROPS IN EMERGING ECONOMIES

Rice area – Philippines, Java, Tamil Nadu, Cambodia, Thailand

Page 11: RIICE service Earth Observation componentsari.umd.edu/sites/default/files/Franceso Holecz.pdf · • Soil data • Phenological data • Management data Yield estimation ... mobile

REMOTE SENSING‐BASED INFORMATION AND INSURANCE FOR CROPS IN EMERGING ECONOMIES

Area accuracy and rice management practices

Crop establishment

method

Number of sites

Avg.Accuracy

Transplanting 6 89.7

Transplanting / Direct Seeding 4 88.5

Direct seeding 3 86.0

Water management

Number of sites

Avg. Accuracy

Irrigated 10 88.8

Irrigated / Rainfed 1 89.0

Rainfed 1 86.0

Semi Dry 1 87.0

Maturity / duration

Number of sites

Avg. Accuracy

Long 4 89.7

Medium 7 88.9

Short 2 85.0

Page 12: RIICE service Earth Observation componentsari.umd.edu/sites/default/files/Franceso Holecz.pdf · • Soil data • Phenological data • Management data Yield estimation ... mobile

REMOTE SENSING‐BASED INFORMATION AND INSURANCE FOR CROPS IN EMERGING ECONOMIES

Yield accuracy and rice management practices

Crop establishment

method

Number of sites

Avg.Accuracy

Transplanting 6 88

Transplanting / Direct Seeding 4 86.5

Direct seeding 3 85.0

Water management

Number of sites

Avg. Accuracy

Irrigated 10 87.8

Irrigated / Rainfed 1 86.1

Rainfed 1 86.5

Semi Dry 1 88.0

Maturity / duration

Number of sites

Avg. Accuracy

Long 4 85.8

Medium 7 87.7

Short 2 90.0

Page 13: RIICE service Earth Observation componentsari.umd.edu/sites/default/files/Franceso Holecz.pdf · • Soil data • Phenological data • Management data Yield estimation ... mobile

REMOTE SENSING‐BASED INFORMATION AND INSURANCE FOR CROPS IN EMERGING ECONOMIES

Sentinel-1A moisac created with MAPscape-RICE © Copernicus data (2015)

Sentinel‐1 – National to continental scale

1. C‐band VV/VH time‐series

2. Coherence time‐series

3. Landsat‐8 time‐series

4. Sentinel‐2 time‐series

Page 14: RIICE service Earth Observation componentsari.umd.edu/sites/default/files/Franceso Holecz.pdf · • Soil data • Phenological data • Management data Yield estimation ... mobile

REMOTE SENSING‐BASED INFORMATION AND INSURANCE FOR CROPS IN EMERGING ECONOMIES

Cambodia – Rice Eco‐system map

Based on Sentinel‐1 12 days VV/VH data acquired from January 2016 to March 2017  

Page 15: RIICE service Earth Observation componentsari.umd.edu/sites/default/files/Franceso Holecz.pdf · • Soil data • Phenological data • Management data Yield estimation ... mobile

REMOTE SENSING‐BASED INFORMATION AND INSURANCE FOR CROPS IN EMERGING ECONOMIES

Cambodia – Dry season 2015‐16

Page 16: RIICE service Earth Observation componentsari.umd.edu/sites/default/files/Franceso Holecz.pdf · • Soil data • Phenological data • Management data Yield estimation ... mobile

REMOTE SENSING‐BASED INFORMATION AND INSURANCE FOR CROPS IN EMERGING ECONOMIES

Cambodia – Early Wet Season 2016 – April Drought

vegetated/water slightly vegetated bare soil/dry veg bare soil

Page 17: RIICE service Earth Observation componentsari.umd.edu/sites/default/files/Franceso Holecz.pdf · • Soil data • Phenological data • Management data Yield estimation ... mobile

REMOTE SENSING‐BASED INFORMATION AND INSURANCE FOR CROPS IN EMERGING ECONOMIES

Mekong River Delta – Winter Spring season 2014‐15 and 2015‐16

Page 18: RIICE service Earth Observation componentsari.umd.edu/sites/default/files/Franceso Holecz.pdf · • Soil data • Phenological data • Management data Yield estimation ... mobile

REMOTE SENSING‐BASED INFORMATION AND INSURANCE FOR CROPS IN EMERGING ECONOMIES

Mekong River Delta – Winter Spring season 2015‐16

Difference of around 7%

0

50,000

100,000

150,000

200,000

250,000

300,000

350,000

An Giang Bac Lieu Ben Tre Ca Mau Can Tho DongThap

HauGiang

Ho ChiMinh

City|HoChi Minh

KienGiang

Long An SocTrang

TienGiang

Tra Vinh VinhLong

Rice Area 2014‐15

Rice Area 2015‐16

Page 19: RIICE service Earth Observation componentsari.umd.edu/sites/default/files/Franceso Holecz.pdf · • Soil data • Phenological data • Management data Yield estimation ... mobile

REMOTE SENSING‐BASED INFORMATION AND INSURANCE FOR CROPS IN EMERGING ECONOMIES

Mekong River Delta – Winter Spring season 2014‐15 vs 2015‐16

0

5

10

15

20

25

30

35

0

5

10

15

20

25

Start of Season – Percentage of rice planted for An Giang province

2014‐15

2015‐16

Page 20: RIICE service Earth Observation componentsari.umd.edu/sites/default/files/Franceso Holecz.pdf · • Soil data • Phenological data • Management data Yield estimation ... mobile

REMOTE SENSING‐BASED INFORMATION AND INSURANCE FOR CROPS IN EMERGING ECONOMIES

Thailand – 2016 Wet Season

1 million ha (18%) were planted too late

Page 21: RIICE service Earth Observation componentsari.umd.edu/sites/default/files/Franceso Holecz.pdf · • Soil data • Phenological data • Management data Yield estimation ... mobile

REMOTE SENSING‐BASED INFORMATION AND INSURANCE FOR CROPS IN EMERGING ECONOMIES

Thailand – 2016 Wet Season

‐25

‐20

‐15

‐10

‐5

010‐May‐16 30‐May‐16 19‐Jun‐16 09‐Jul‐16 29‐Jul‐16 18‐Aug‐16 07‐Sep‐16 27‐Sep‐16 17‐Oct‐16

AYTPT1BPH_07

AYTPT1BPH_08

AYTPT1BPH_09

AYTPT1BPH_10

AYTPT1BPH_11

Seasonal monitoring  from satellite

Anomalies = disease and lodging

3D drone image

Page 22: RIICE service Earth Observation componentsari.umd.edu/sites/default/files/Franceso Holecz.pdf · • Soil data • Phenological data • Management data Yield estimation ... mobile

REMOTE SENSING‐BASED INFORMATION AND INSURANCE FOR CROPS IN EMERGING ECONOMIES

Tamil Nadu – Samba season 2015‐16

Page 23: RIICE service Earth Observation componentsari.umd.edu/sites/default/files/Franceso Holecz.pdf · • Soil data • Phenological data • Management data Yield estimation ... mobile

REMOTE SENSING‐BASED INFORMATION AND INSURANCE FOR CROPS IN EMERGING ECONOMIES

Page 24: RIICE service Earth Observation componentsari.umd.edu/sites/default/files/Franceso Holecz.pdf · • Soil data • Phenological data • Management data Yield estimation ... mobile

REMOTE SENSING‐BASED INFORMATION AND INSURANCE FOR CROPS IN EMERGING ECONOMIES

Field work and validation

Fieldwork, based on well established and consistent field protocols, is crucial:

• To understand rice practices (‘data calibration’)

• To validate the products

• To obtain information for improvements and extending them

• To alert to any problems

Classification RICEClassificationNOT RICE Producer’s Accuracy

Reference RICE 1639 194 89.4

Reference NOT RICE 157 1237 88.7

User’s accuracy 91.3 86.4

Page 25: RIICE service Earth Observation componentsari.umd.edu/sites/default/files/Franceso Holecz.pdf · • Soil data • Phenological data • Management data Yield estimation ... mobile

REMOTE SENSING‐BASED INFORMATION AND INSURANCE FOR CROPS IN EMERGING ECONOMIES

Capacity building

• Capacity building events in each country every year.

• Capacity building on remote sensing, crop modeling and field activities.

• A total of 50 training courses have been carried out.

• The RIICE service (MAPscape‐RICE, Oryza, and field protocol) has all improved

thanks to feedback from national partners.

Page 26: RIICE service Earth Observation componentsari.umd.edu/sites/default/files/Franceso Holecz.pdf · • Soil data • Phenological data • Management data Yield estimation ... mobile

REMOTE SENSING‐BASED INFORMATION AND INSURANCE FOR CROPS IN EMERGING ECONOMIES

Bulletins

Page 27: RIICE service Earth Observation componentsari.umd.edu/sites/default/files/Franceso Holecz.pdf · • Soil data • Phenological data • Management data Yield estimation ... mobile

REMOTE SENSING‐BASED INFORMATION AND INSURANCE FOR CROPS IN EMERGING ECONOMIES

Philippines Rice Information SysteM – PRISM 

Page 28: RIICE service Earth Observation componentsari.umd.edu/sites/default/files/Franceso Holecz.pdf · • Soil data • Phenological data • Management data Yield estimation ... mobile

REMOTE SENSING‐BASED INFORMATION AND INSURANCE FOR CROPS IN EMERGING ECONOMIES

Where today RIICE service is used

• Philippines – The Philippine Rice Information SysteM (PRISM)

• Vietnam

• Cambodia

• Thailand

• India, Tamil Nadu

• India, Odisha

• India, Andhra Pradesh

• India, other states have shown interest

Page 29: RIICE service Earth Observation componentsari.umd.edu/sites/default/files/Franceso Holecz.pdf · • Soil data • Phenological data • Management data Yield estimation ... mobile

REMOTE SENSING‐BASED INFORMATION AND INSURANCE FOR CROPS IN EMERGING ECONOMIES

Conclusions

• Understanding of crop practices and fieldwork calibration/validation are bothconditio sine qua non (learning and credibility).

• The availability of systematic multi‐sensor acquisitions is essential for anoperational service in particular for agricultural applications (complementarityand redundancy).

• The use of spatial data in crop yield modeling, so far limited to point data, iscrucial.

• It is essential that national partners have an active role, in particular wrt: local expertise access to field sample data for calibration/validation products acceptance drivers of education in the country

Page 30: RIICE service Earth Observation componentsari.umd.edu/sites/default/files/Franceso Holecz.pdf · • Soil data • Phenological data • Management data Yield estimation ... mobile

REMOTE SENSING‐BASED INFORMATION AND INSURANCE FOR CROPS IN EMERGING ECONOMIES

Thank you for your attention