1. Introduction
Total population of the Kenya at 2014
was 45.5 Million individuals that the share of
rural and urban area was 74.4% and 25.6%
respectively. Within the crop income category,
maize with horticultural crops, covering 14%
of total household income in Kenya. At the
same time, coffee and tea account for a
combined 5.6% of total gross income. In the
mid-1970s maize accounted for about 35% of
the value of total crop production in Kenya
while in the late 1990s, it reduced to 28% of
the value of total crop production (Olwande,
Sikei et al. 2009). Since 2008 the area under
cereal production increased ~ 647,762 ha and
the cereal production reached 1,446,328
tonnes.
Maize farming is the backbone of food
security in Kenya. Maize fields increased from
1,700,000 ha at 2008 to 2,159,322 ha at 2012,
which is the highest since 1961. Amid 2008
and 2012, maize production increased
1,382,643 tonnes. At 2014 around Kenya
harvested 3,513,171 tonnes maize although it
An Investigation into Maize Lethal Necrosis Severity Mapping in
Kenya using RapidEye and Landsat-8
Hosein Jafari Jozani (H.J.), Michael Thiel (M.T.), Michael Hahn (M.H.), Kyalo Richard
(K.R.), and Tobias Landmann (T.L.)
Department of Remote Sensing, University of Würzburg, Oswald-Külpe-Weg 86, 97074
Würzburg, Germany; michael.thiel@uni-wuerzburg.de (M.T.)
Department of Geomatics, University of Applied Science Stuttgart (HFT), Schellingstraße 24,
70174 Stuttgart, Germany; michael.hahn@hft-stuttgart.de (M.H.); hoseinjafary@aol.com (H.J.)
International Center for Insect Physiology and Ecology (ICIPE), P.O. Box 30772, 00100
Nairobi, Kenya; rkyalo@icipe.org (K.R.);
Center for Development Research (ZEF), Department of Ecology and Natural Resources
Management, University of Bonn, Walter-Flex-Str. 3, 53113 Bonn, Germany; cb@uni-bonn.de
Abstract
The spatial information of crops and cropping systems can provide useful information about
disease outbreak mechanism in croplands. In September 2011, a severe outbreak of Maize Lethal
Necrosis (MLN) reported at lower elevations (1,900 meter above sea level) in Southern Rift
Valley, Longisa division of Bomet County, Kenya. The heterogeneous small scale farms and
cloud cover in study area are some of challenges against application of remote sensing in this
region. Aims of this study are; to classify maize fields, to discriminate mono/intercropping,
continuous/rotation cropping systems, to identify severe MLN occurrence and finally to
determine if any relevance between severe MLN occurrences with cropping system and
ecological variables exist. Infield data collection was accomplished by African insect for food
and health (icipe) (December 2014, January and August 2015), which provided GPS data, crop
type, crop conditions (Type, Physical condition, growth stage, and MLN severity), cropping
system (mono/inter cropping) and crop rotation (continuous/rotation). High-resolution RapidEye
(RE) images, medium resolution Synthetic Aperture RADAR (SAR) Sentinel-1 (S1), Shuttle
Radar Topography Mission (SRTM) Digital Elevation Model (DEM) of 30 meter resolution and
12-monthes of Landsat-8 (L8) images were acquired. Several vegetation indices were calculated
from RE and L8 adopted for classification. The Random Forest (RF) and One Class Classifier
(OCC) were performed in the absence of complete and representative validation data of Land
Cover Land Use (LCLU) samples, and the results compared. The RE and L8 images covered
two cropping seasons. Also Normalized Difference Vegetation Index (NDVI) in regional-level
time series calculated out of 12 L8 images to derive and analysed. As first step, RF classifier
employed to produce a LULC map. Secondly, mono and intercropping maize fields identified via
maize fields reclassification. The third step, severe MLN classified via adopting OCC classifier.
As fourth step, cropping system determined (continuous/rotational) through applying raster
analysis. Finally the relationship among cropping system, ecological variables and MLN severity
was investigated if any.
Key words: Cropping systems, RF, OCC, MLN severity, NDVI, Ecological variables.
Preprints (www.preprints.org) | NOT PEER-REVIEWED | Posted: 4 May 2018 doi:10.20944/preprints201805.0095.v1
© 2018 by the author(s). Distributed under a Creative Commons CC BY license.