Print This Page

JECAM | Joint Experiment for Crop Assessment and Monitoring

South Africa

Project Overview

Crop identification and Crop Area Estimation

The Umlindi system, based on course resolution imagery (Proba V and Modis) and in situ (weather station data) continues to issue its monthly newsletter on drought monitoring, crop condition and other relevant information. The system is continually improved as new indices are incorporated.

The Umlindi system issues a monthly soil water product generated using the Topkapi model run by the University of KwaZulu-Natal. A number of soil water sensors have also been added to some of the automatic weather stations located in the JECAM site.

Yield data for wheat and maize is also operationally collected and all this data is fed to the national Crop Estimations Committee for monthly national production estimates. Operational systems are continually improved as technologies and systems become available.

Not currently addressed due to resource constraints.

The mapping resolution is around 1:50 000. 

The timeliness (with regards to growing season) are:


Project Reports

2016 Site Progress Report

2015 Site Progress Report

2014 Site Progress Report

 

Yacob G. Beletse, Wiltrud Durand, Charles Nhemachena, Olivier Crespo, Weldemichael A. Tesfuhuney, Matthew R. Jones, Mogos Y. Teweldemedhin, Sunshine M. Gamedze, Pontsho M. Bonolo, Syanda Jonas, Sue Walker, Patrick Gwimbi, Thembeka N. Mpuisang, Davide Cammarano, Roberto O. Valdivia. 2015, Chapter 4: Projected Impacts of Climate Change Scenarios on the Production of Maize in Southern Africa: An Integrated Assessment Case Study of the Bethlehem District, Central Free State, South Africa. In Rosenzweig C, & D Hillel (2015) The Agricultural Model Intercomparison and Improvement Project (AgMIP) Integrated Crop and Economic Assessments — Joint Publication with American Society of Agronomy, Crop Science Society of America, and Soil Science Society of America (In 2 Parts).

Sinclair S. and Pegram G.G.S., (2013), "A sensitivity assessment of the TOPKAPI model with an added infiltration module", J. Hydrol., 479, 100-112.

Sinclair S. and Pegram G.G.S., (2013), HYLARSMET: A Hydrologically Consistent Land Surface Model for Soil Moisture and Evapotranspiration Modelling over Southern Africa using Remote Sensing and Meteorological Data, WRC Report No. 2024/1/13, Water Research Commission, Pretoria, South Africa.

Tsoeleng, L., Knox, N., Mashalane, M., Sibandze, P., Mangara, P., Kganyago, M., Oliphant, T., Mahlangu, N., Mudau, N., Mapurisa, W., Meyer, B., Agricultural Crop Identification. Oral Presentation: SA-GEO Conference, University of Fort Hare, 2013.


Implementation Plans

n Situ Data:

Climate Data

Forty-two weather stations within the JECAM site collect data daily. Parameters include rainfall, temperature, wind speed and direction, and soil water at a selected number of stations (5). The following graph shows data collected at one of the stations for rainfall and soil moisture.

Figure 1 Hourly Rainfall (bars) and Soil Moisture (% at 10 cm and 50 cm) at Bethlehem for the Period 14 January to 14 February 2016

Crop Type Observations

The Producer Independent Crop Estimation System (PICES) collects around 4500 visual observations of crop type with in fields (and some ancillary data) from light aircraft for the winter and summer growing season. These points comprise observations of statistically selected sample points as well as opportunistic observations between sample points while traversing the agricultural area. This data is used to estimate crop type areas.

 

Yield Measurements

At around 670 points, yield measurements are collected in situ for wheat and maize. These are used to estimate average yield for the province for the two crops.

 

Figure 2 shows the crop condition on a scale of 1 (good) to 9 (total failure) on 18 February 2015. Drought ravaged some areas of the Free State.

Figure 2 Crop Condition at 18 February 2105

Plans for Next Growing Season:

We plan to maintain the current approach next year. The EO data to be ordered next year will include Sentinel-2 data when available.


Site Description

Locations

South Africa Western Freestate (Sub-site for Intensive Observation 2)
Site Extent   Centroid: -28.144, 26.157
Top left: -28.029, 26.026 Bottom Right: -28.260, 26.289

South Africa Full Site
Site Extent   Centroid: -28.428, 27.065
Top left: -26.851, 24.548 Bottom Right: -30.741, 29.770

South Africa Eastern Freestate (Sub-site for Intensive Observation 1)
Site Extent   Centroid: -28.113, 28.694
Top left: -28.002, 28.564 Bottom Right: -28.225, 28.823

Typical field size range: 0.5 ha to 40 Ha.  Crop types are Winter – Wheat and Oats and  Summer – Maize, Sunflower, Soya, Groundnuts, Sorghum, Dry Beans.  The crop rotation patterns vary from maize followed by legume (soy beans, groundnuts or dry beans), or maize followed by fallow the next summer with wheat in the immediately following winter. Sunflower is sometimes interchanged with Maize as a summer crop depending on the current prices at the time of planting. The majority of grain that is irrigated is under centre pivot systems, and in many cases it is a double cropping rotation with winter wheat followed immediately with summer maize.

The climatic zone is sub-humid to semi-arid. (Summer convectional rainfall)

The total cropping season covers an entire calendar:  1 July 2012 up to 30 June 2013 (12 months) The Free State province has a continuous cropping season that include both summer crops (maize, sunflower, soy beans, sorghum, groundnuts etc) as well as winter crops (wheat, oats etc). and spatially distributed across the province.

Typical Crop calendar (Colours in lower rows simulate CIR imagery)

Therefore there is no specific end of season or fixed area for summer vs winter crops. Lucern (also known as Alfalfa) is a perennial crop that is grown under irrigation, of which the majority is centre pivot irrigation.

Soil Texture
Cultivation occurs on sandy to sandy-loam soils. Wind erosion is often an issue

Landscape Topology
Landscape ranges from flat undulating plains to mountainous in the east.

Topography of the Free State Province

Soil Drainage
Soil drainage ranges from well drained areas to areas with very high water tables.

Soil Drainage Classes of Free State Province

Other Site Specifications
Irrigation infrastructure is mainly centre pivots

Topo-cadastral map of the Free State Province showing Roads and towns

Land Cover Map of the Free State Province

Climate Zones of the Free State Province

Distribution of automatic weather stations


Specific Project Objectives & Deliverables

Results

The crop monitoring and estimation system operational in this JECAM site provided data to the National Crop Estimation Committee. This committee issued the following official estimates for the Province:

Wheat (2015 season):  80,000Ha planted, 208,000 tons harvested (6th production forecast – 27 Jan 2016)

Total Maize (2014/2015 season):       1,220,000 Ha planted, 3,945,000 tons harvested (Final production forecast 29 Sept 2015)

Yellow Maize (2014/2015 season):  510,000 Ha planted, 1,708,500 tons harvested (Final production forecast 29 Sept 2015)

White Maize (2014/2015 season): 710,000 Ha planted, 2,236,500 tons harvested (Final production forecast 29 Sept 2015)

Sunflower:    285,000 Ha planted, 370,500 Tons harvested (Final production forecast 29 Sept 2015)

Sorghum: 36,000Ha planted, 45,000 tons harvested (Final production forecast 29 Sept 2015)

The operational system is working well and is being continually improved as research results are operationalized.

The current operational system, PICES, and the yield estimation system that supplies crop estimate information to the national Crop Estimates Committee are considered a best practice methodology.

This is confirmed by a recent quote from the Crop Estimates Liaison Committee representing the South African Agricultural grain crop industry. “The Crop Estimates Liaison Committee, as oversight body over the Crop Estimates Committee, congratulates and thanks the CEC for the sterling work they did over this past season. The 2015 season was exceptionally difficult since a serious drought and heat wave made crop estimates extremely difficult, but the CEC nevertheless were well within the 5% deviation target with their maize estimates right from the first estimate in February 2015. That is exceptional and needs to be recognized”, Dr John Purchase, Chairman of the CELC (and member of the GEOGLAM Steering Committee) said today (11 Feb 2016).


In Situ Observations

  1. Parameter: Yield measurements
    Data Collection Protocol:

    Routinely collected at a random number of points

    Frequency: 150 summer, 70 Winter
  2. Parameter: Temperature, Humidity, Rainfall, Solar Radiation, Wind, Potential evapotranspiration (derived)
    Data Collection Protocol:

    Weather station data from the automatic weather station network of the ARC-ISCW have been collected at/near each of the 2 sub-sites.

    Frequency:

EO Data Requirements

Approximate Start Date of Acquisition: August 2012 (min) - July 2012 (preferred): Need to assess onset of rainfall, planting and area planted. Summer crops are planted as early as October in certain parts of the Free State
Approximate End Date of Acquisition: November 2013: Need also to assess winter grain over the Free State, harvested around November
Spatial Resolution: 2.5 m (min) - 1 km (preferred)
Temporal Frequency: 10-Daily: Need to build a phonological time series from multi-temporal data
Latency of Data Delivery: 5 days
Wavelengths Required: Red, NIR, Blue: Needed for sensing vegetation activity and vegetation type
Across Swath:
Along Track:

SAR Data Requirements

Approximate Start Date of Acquisition: August 2012 (min) - July 2012 (preferred)
Approximate End Date of Acquisition: May 2013 (min) - June 2013 (preferred)
Spatial Resolution:
Temporal Frequency: 2 monthly (min) - 1 monthly (preferred)
Latency of Data Delivery:
Wavelengths Required: C-Band & X-Band
Polarization VV
Incidence Angle Restrictions:
Across Track:
Along Track:

Locations

South Africa Western Freestate (Sub-site for Intensive Observation 2)

Centroid
Latitude: -28.144
Longitude: 26.157

Site Extent
Top left
Latitude: -28.029
Longitude: 26.157
Bottom Right
Latitude: -28.260
Longitude: 26.289

South Africa Full Site

Centroid
Latitude: -28.428
Longitude: 27.065

Site Extent
Top left
Latitude: -26.851
Longitude: 27.065
Bottom Right
Latitude: -30.741
Longitude: 29.770

South Africa Eastern Freestate (Sub-site for Intensive Observation 1)

Centroid
Latitude: -28.113
Longitude: 28.694

Site Extent
Top left
Latitude: -28.002
Longitude: 28.694
Bottom Right
Latitude: -28.225
Longitude: 28.823


Optical Sensors

SPOT VEGETATION
Imaging Mode:
Spatial Resolution: 1 km
Acquisition Frequency: 10-daily
Pre-Processing Level:
Application: Crop monitoring and modelling

MODIS
Imaging Mode: 16-day
Spatial Resolution: 250-1000m
Acquisition Frequency: 4 per study area
Pre-Processing Level: Q1
Application:

SPOT 5
Imaging Mode: Multispectral (10 m) and Panchromatic (2.5m)
Spatial Resolution: 10/2.5m
Acquisition Frequency: 01 Jan 2013 – 31 Dec 2014 (National Mosaic)
Pre-Processing Level: Ortho-rectified
Application: Area planted, crop type, field boundaries

Proba-V
Imaging Mode:
Spatial Resolution: 1km
Acquisition Frequency: daily
Pre-Processing Level:
Application:

ResourceSat Awifs
Imaging Mode: Awifs
Spatial Resolution: 50m
Acquisition Frequency: Monthly
Pre-Processing Level: Ortho-rectified
Application: Crop type classifications for area estimates

Landsat-8-OLI
Imaging Mode:
Spatial Resolution:
Acquisition Frequency: 3 times, Apr-Jun 2013
Pre-Processing Level:
Application: crop type identification

Rapid Eye (sub-sites 1 & 2 only)
Imaging Mode: Multispectral
Spatial Resolution: 5m
Acquisition Frequency: Monthly
Pre-Processing Level: Ortho-rectified
Application: Crop type classifications for area estimates

SPOT-4 Take 5
Imaging Mode:
Spatial Resolution:
Acquisition Frequency: 17 cloud-free
Pre-Processing Level: 2A
Application:

WorldView 2 (sub-sites 1 & 2 only)
Imaging Mode: Multispectral
Spatial Resolution: 2m
Acquisition Frequency: Monthly
Pre-Processing Level: Ortho-rectified
Application: Crop type classifications for area estimates

JECAM | Joint Experiment for Crop Assessment and Monitoring | Group on Earth Observation

©2013 Joint Experiment for Crop Assessment and Monitoring © HER MAJESTY THE QUEEN IN RIGHT OF CANADA SA MAJESTE LA REINE DU CHEF DU CANADA (2012)