Abstract
Rice is the most important food crop in Asia and the mapping and monitoring of paddy
rice fields is an important task in the context of food security, food trade policy and
greenhouse gas emissions modelling. Two countries where rice is of special
significance are China, the largest producer and importer of rice, and Vietnam, where
rice exports contribute a fifth to the GDP. Both countries are facing increasing pressure
in terms of food security due to population and economic growth while agricultural
areas are confronted with urban encroachment and the limits of yield increase.
Despite the importance of knowledge about rice production the countries official land
cover products and rice production statistics are of varying quality and sometimes even
contradict each other. Available remote sensing studies focused either on time-series
analysis from optical sensors or from Synthetic Aperture Radar (SAR) sensors ? the
studies using optical sensors faced problems due to either the spatial or temporal
resolution and the persistent cloud cover while SAR studies found the limited data
availability and large image size to be the biggest drawbacks. We try to address these
issues by proposing a paddy rice mapping approach that combines medium spatial
resolution, temporally dense time-series from the optical MODIS sensors and high
spatial resolution time-series from the recently launched Sentinel-1 SAR sensor.
We used the 250m resolution MOD13Q1 and MYD13Q1 products as a basis for our
medium resolution rice map. Prevalent cloud cover introduces noise into these timeseries
which we reduced by applying a Savitzky-Golay filter. We then derived a number
of time-series temporal and phenological metrics for multiple years and classified rice
areas with One Class Support Vector Machines. In a next step we used this medium
resolution rice map to mask Sentinel-1 Interferometric Wide Swath images and create
SAR time-series from which we again derived temporal and phenological metrics and
classified rice areas with machine learning algorithms to arrive at a 10m resolution rice
map.
This method allows concurrent, accurate and high resolution mapping of paddy rice
areas from freely available data with limited requirements towards processing
infrastructure and can be used as a basis for greenhouse gas and crop modelling as
well as providing viable information for decision makers regarding food security, food
trade, bioeconomy and mitigation after crop failure. Results of our paddy rice
classification will be presented for selected study sites in China and Vietnam.
Users
Please
log in to take part in the discussion (add own reviews or comments).