Access to this full-text is provided by Taylor & Francis.
Content available from GIScience & Remote Sensing
This content is subject to copyright. Terms and conditions apply.
Earth observations reveal impacts of climate variability on maize cropping
systems in sub-Saharan Africa
Adomas Liepa
a
, Michael Thiel
a
, Hannes Taubenböck
b,c
, Doris Klein
b,c
, Ingolf Stean-Dewenter
d
,
Marcell K. Peters
d
, Sarah Schönbrodt-Stitt
a
, Insa Otte
a
, Tobias Landmann
e
, Zeyaur R. Khan
e
,
Michael Ochieng Obondo
e
, Frank Chidawanyika
e,f
, Emily A. Martin
g
and Tobias Ullmann
a
a
Earth Observation Research Cluster, Department of Remote Sensing, Institute of Geography and Geology, University of Würzburg, Würzburg,
Germany;
b
Earth Observation Research Cluster, Department of Global Urbanization and Remote Sensing, Institute of Geography and Geology,
University Würzburg, Würzburg, Germany;
c
German Aerospace Center (DLR), German Remote Sensing Data Center (DFD), Oberpfaffenhofen,
Germany;
d
Department of Animal Ecology and Tropical Biology Biocentre, University of Würzburg, Würzburg, Germany;
e
Geo-Information
Unit, International Center for Insect Physiology and Ecology (ICIPE), Nairobi, Kenya;
f
Department of Zoology and Entomology, University of the
Free State, Bloemfontein, Republic of South Africa;
g
Department of Animal Ecology & Systemics, Justus Liebig University of Gießen, Gießen,
Germany
ABSTRACT
In Kenya, climate variability and change threaten smallholder, rainfed farms, with crop fail-
ures, yield reductions, and pest infestations. Ecient agroecological strategies, such as Push-
Pull intercropping, oer documented benets including pest control, improved soil fertility,
and water conservation compared to traditional maize monocropping. To date, no studies
exist comparing traditional maize monocropping and Push-Pull intercropping using earth
observation tools over several growing seasons in East Africa. Our research addresses this by
harmonizing Landsat 7, 8, 9 with Sentinel-2 remote sensing time series from 2016 to 2023.
Phenological metrics of 15 growing seasons are extracted based on a threshold method
using the Normalized Dierence Vegetation Index (NDVI) as a vegetation proxy. Field data
from 58 sites in southwestern Kenya provided training for this analysis, revealing detectable
inter-class dierences. Notably, Push-Pull intercrop elds showed greater resilience during
biotic stress events, such as the locust outbreak in 2020 short rainy season and the fall
armyworm infestation in combination with delayed and below-average rainfall during the
short 2021 and the long 2022 growing seasons. Higher maximum NDVI and extended season
duration indicated a higher resilience of Push-Pull farming under unfavorable agricultural
conditions. Short growing seasons with unfavorable conditions showed earlier end of seasons
in both systems, whereas long growing seasons with unfavorable conditions caused delayed
onset and end of seasons. This study marks the rst attempt to leverage earth observation
data to compare traditional maize agriculture with agricultural systems featuring applied
ecological management strategies, showcasing the potential of earth observation tools to
monitor and evaluate agroecological resilience.
ARTICLE HISTORY
Received 15 July 2024
Accepted 28 February 2025
KEYWORDS
Agriculture; phenology;
Africa; earth observation;
agroecological practice;
remote sensing
CONTACT Adomas Liepa adomas.liepa@uni-wuerzburg.de
Supplemental data for this article can be accessed online at https://doi.org/10.1080/15481603.2025.2476248
GISCIENCE & REMOTE SENSING
2025, VOL. 62, NO. 1, 2476248
https://doi.org/10.1080/15481603.2025.2476248
© 2025 The Author(s). Published by Informa UK Limited, trading as Taylor & Francis Group.
This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use,
distribution, and reproduction in any medium, provided the original work is properly cited. The terms on which this article has been published allow the posting of the Accepted
Manuscript in a repository by the author(s) or with their consent.
1. Introduction
Climate change can be dened as signicant long-
term variations in meteorological conditions such as
precipitation and temperature (Allen 2003; World
Meteorological Organization 1992). The eects of
changing climate are inevitable and felt globally
(Adams et al. 1998; Aydinalp and Cresser 2008;
Salvo, Begalli, and Signorello 2013). It is a cause of
concern considering the importance of variations in
meteorological parameters on agricultural production
(Sharma et al. 2022; Torres, Howitt, and Rodrigues
2019). This becomes more alarming considering
that, to meet the food demand of the projected glo-
bal population of 9 billion by 2050, an increase in
staple cereal production of +70% is required
(Neupane et al. 2022).
The gravity of climate change impact on agricul-
tural yield varies by crop type, geographical location
and local climate (Challinor et al. 2014). The impact of
climate change is distributed unevenly worldwide.
The current climatic tendencies will lead to destruc-
tive changes to sub-Saharan Africa (Mubenga-
Tshitaka et al. 2023; Schlenker and Lobell 2010). In
East Africa, current trends indicate up to 20% increase
in rainfall from December to February and a decrease
in rainfall of up to 10% from June to August accom-
panied by higher projected temperatures of up to
1.9°C by the year 2050 (Gebrechorkos, Hülsmann,
and Bernhofer 2019; Hulme et al. 2001; IPCC 2023).
In addition, highly dynamic population growth (UN
2018) and a very dynamic urbanization trend
(Taubenböck et al. 2024) predicted for sub-Saharan
Africa will increase food insecurity.
East African countries are highly reliant on the
agricultural sector. For Kenya, agriculture accounts
for an estimated 21.2% of Gross Domestic Product
(GDP) and employs over 70% of the rural population
(The World Bank 2022). Moreover, smallholder farm-
ers, most vulnerable demographic, account for more
than 80% of total agricultural output of Kenya (Market
Alliance 2022).
Changing climate results in nightly and daily warm-
ing as well as shifts in precipitation patterns in Kenya
(Malhi, Kaur, and Kaushik 2021; Torres, Howitt, and
Rodrigues 2019). Temperature and precipitation are
direct inputs in agricultural production, hence any
changes to these parameters will aect the agricul-
tural output (Deschenes and Greenstone 2004), i.e.
through abiotic stresses (Halford et al. 2015; Sánchez-
Bermúdez et al. 2022). Based on Kenya’s annual maize
yield data from 1979 to 2012, signicant increases in
temperature and reduction in seasonal rainfall
resulted in maize yield decreases of 0.07 tons/ha/
decade (Mumo, Yu, and Fang 2018). More recent
maize yield data from 2012 to 2022 have shown
increases in maize yield during growing seasons
with increased seasonal rainfall (Ondiek, Saber, and
Abdel-Fattah 2024). Research using household sur-
veys mirror the ndings of numerical and statistical
studies. Ochieng, Kirimi, and Mathenge (2016) found
that long-term temperature increase has a larger
impact on small-scale crop production than short-
term precipitation uctuations. Severe temperature
during the vegetative state has also led to decreased
cereal kernel quality and amounts (Hütsch, Jahn, and
Schubert 2019). Timely rainfall can be a mitigating
factor; however, severe uctuations in precipitation
can also lead to outstanding yield loss and crop failure
(GEOGLAM 2019). To make matters worse, negative
eects on agriculture through biotic stresses such as
presence of weeds, pest outbreaks and soil fertility
decrease have also been widely recognized (Jafari
Jozani et al. 2022; Shahzad et al. 2021). Above-
average rains in 2019 led to wet soil and vegetation
conditions which subsequently created favorable set-
tings for locust gregarization in East-Africa (Cressman
2013). The fall armyworm (FAW) invasion since
2016 has caused severe damage to maize yield and
is a major threat to food security throughout East
Africa (Sisay et al. 2019; Tambo et al. 2020). Other
important constraints to maize production include
the lepidopterous stemborers (Kr et al. 2002;
Midega et al. 2015) and parasitic Striga weed (David
et al. 2022; Khan et al. 2002), which can cause mea-
sured maize yield losses of up to 88%.
Considering these challenges, agroecological man-
agement strategies for the control of stemborers and
striga weeds in maize elds have been widely devel-
oped in providing a biological alternative to expen-
sive and harmful pesticides. By 2016, Push-Pull
intercropping have been employed by over 125.000
smallholder farmers in East Africa leading to substan-
tial maize yield increases (Khan et al. 2016). Push-Pull
is a stimulo-deterrent diversionary strategy that relies
on behavioral manipulation using airborne volatile
2A. LIEPA ET AL.
organic compound from companion plants to man-
age crop pests. To achieve this, the main cereal crop,
such as maize, is intercropped with a repellent plant
Desmodium sp., that pushes the stemborer and FAW
pests from the main crop. The Desmodium root exu-
dates are also known to suppress parasitic Striga
weeds by triggering suicidal germination (Hooper
et al. 2015). In addition, the leguminous Desmodium
xes nitrogen in the soil enhancing soil fertility and
retaining soil moisture in dry conditions. Around the
plot edges, a trap plant, Brachiaria sp., is planted that
naturally attracts stemborer and acts as a pull for the
egg-laying female stemborer (Khan et al. 2016; Khan
et al. 2001; Midega et al. 2015). The push and pull
plants are both valuable additives, as Desmodium is
rich in protein and Brachiaria contains high levels of
carbohydrates and are known to increase milk pro-
duction in cattle. Acting together, these plants pro-
vide a push and pull eect which increases maize
yield, maize resistivity to stemborer, FAW and striga
weeds, enhances soil fertility and promotes water
retention (Buleti et al. 2023). Ecological management
such as Push-Pull intercropping has shown promise
for the sustainable management of insect pests
(Stean-Dewenter, Kerr, and Rachel 2024). Yet, few
studies have compared the response of Push-Pull
intercrops to external forcings with that of traditional
maize monocrop elds on a temporal scale using
earth observation (EO) data.
The smallholder farms in Kenya are largely rainfed,
making them highly dependent on timely seasonal
rains (Richard et al. 2017). It is common to see the
yearly total precipitation relatively unchanged, while
the timing of the rainfall and its intensity become
steadily more dicult to predict and adapt to (Otte
et al. 2017; Torres, Howitt, and Rodrigues 2019). The
onset of the seasonal rainfall is an important factor
contributing to a successful growing season. The tim-
ing of the rains will determine the planning and pre-
paration of the land as well as the sowing of the crop
(Ojo, Temenu, and Ilunga 2019). Therefore, the tem-
poral variations in rainfall and its intensity have
a direct impact on food supply.
In this regard, capturing the seasonal phenology of
agriculture is crucial to provide timely information on
plant responses to external forcings. Despite the added
value of EO already demonstrated in a range of agri-
cultural applications (Gao and Zhang 2021; Worrall
et al. 2023), many studies rely on traditional crop
monitoring through household surveys, which leads
to sparse information based on little to no integration
of EO-based crop yield and condition models
(Nakalembe et al. 2021, Qader et al. 2021). This
approach suers from several limitations: (i) it requires
time and a high amount of human resources, (ii) the
generated output provides only a local assessment
preventing the results from being generalized across
relevant regional scales, and (iii) the data collection at
a single timestep prevents incorporation of a temporal
element (Henrys and Jarvis 2019). In contrast, earth
observation presents a signicant potential to over-
come such limitations by introducing remote sensing
data at multiple temporal and spatial resolutions.
The present study focuses on demonstrating the
response of varying management strategies of agri-
culture to external climatic forcings by using the mul-
tispectral-normalized dierence vegetation index
(NDVI) to estimate phenological metrics, such as
start, peak and end of seasons on eld-level. The
method was applied on maize monocrop and maize
Push-Pull intercrop elds in the Lake Victoria region of
Kenya over several growing seasons by employing
a harmonized dense time series of Landsat and
Sentinel-2 observations and ground truth data with
the aim of illustrating the ability of both systems to
perform under varying climatic conditions. We speci-
cally aim to: (i) synthesize available climatological
knowledge to derive an accurate depiction of the
climatic conditions and biotic stresses in the area
from 2016 to 2023, (ii) extract the phenological
metrics of maize monocrop and Push-Pull elds and
(iii) examine the impact of climatic and biotic factors
on the performance of these two agricultural systems.
2. Materials and methods
2.1. Study area
The study area is located in southwestern Kenya
(Figure 1). In terms of climate, the region is part of
the tropical rainforest climate zone and is character-
ized by relatively constant high temperatures of 18°C
or higher and generally high yearly rainfall with over
2.000 mm per year (Beck et al. 2018). The climate in
the area is heavily inuenced by the Inter Tropical
Convergence Zone (ITCZ) (Palmer et al. 2023). The
annual changes of the ITCZ result in two wet seasons:
April to June known as the “long rainy season” and
GISCIENCE & REMOTE SENSING 3
October to December referred to as “short rainy sea-
son” (The World Bank 2020). This results in a bimodal
growing season pattern with the growing seasons
occurring in the periods of March to July and
October to December.
The area is characterized by fertile agricultural land
which is predominantly rainfed. As a result, the majority
of the population performs small-scale farming, graz-
ing, and shing. A wide variety of staple and cash crops
are cultivated in the area, most notably maize, cassava,
sweet potato, beans, groundnuts, banana, sugarcane,
sorghum, and coee (Ekesa et al. 2015).
2.2. Data
2.2.1. Satellite data
For the temporal analysis, we used freely available
high-spatial resolution products from USGS Landsat
missions and Copernicus Sentinel-2 program
(Table 2). We used the Landsat Level 2 products,
which contain atmospherically corrected and orthor-
ectied surface reectance values. The Landsat pro-
ducts have a spatial resolution of 30 meters with
a panchromatic band in 15-meter resolution and pro-
vide seven spectral bands of which red and near-
infrared were used in this study to derive the vegeta-
tion index. The Landsat constellation is set up with an
oset which allows a repeat coverage of 8 days, while
the Sentinel-2 mission provides an image scene every
5–6 days. Sentinel-2 sensor provides 12 spectral
bands in 10-to-20-meter spatial resolution. The
Sentinel-2 Level 2 images are atmospherically cor-
rected using ESA´s sen2cor algorithm (Main-Knorn
et al. 2017).
2.2.2. Ground reference and in situ data
A data gathering campaign took place during the
long growing season in May 2023. The campaign
Figure 1. Overview of the study area. Spatial distribution of the in-situ data gathered during the field work is shown in A) where single
points denote fields containing one or several sampled fields, regional extent of the study area in B), and continental extent of the
regional map shown in C). Sentinel-2 cloud free composite is used as a background layer for A) and Natural Earth vector and raster
map data available at naturalearthdata.com/ is used as background for B) and C).
4A. LIEPA ET AL.
focused on surveying the four maize-producing
regions in Kenya close to Lake Victoria: Siaya,
Kisumu, Homa Bay and Migori (Figure 1). During
the survey, corner and center coordinates of Push-
Pull intercrop and maize monocrop elds were col-
lected using GPS Coordinates App Version 4.71
(174). Afterwards the in-situ data was imported to
the QGIS 3.22.6 and with the help of the satellite
basemap, small eld boundary adjustments were
made. We used only elds containing the same
crop type throughout the study period, relying on
information provided by the farmers for pre-
seasons. In total, 58 ground-referenced elds have
been selected of which 26 were Push-Pull and 32
were maize monocrop elds. An example of Push-
Pull intercrop and maize monocrop elds can be
seen in Appendix B. A summary of the eld dimen-
sions for each group can be seen in Table 1. The
farmers were kept in regular contact throughout the
study duration. Field visits by the local agricultural
scientists from International Centre of Insect
Physiology and Ecology (icipe) took place at
a regular basis ensuring continues in situ data vali-
dation. The data gathering campaign concluded
with the interviews of farmers whose elds were
sampled during which the crop type of previous
seasons was conrmed. The farmer interviews
added insights into agricultural tendencies in the
area and individual perspectives on the conditions
of the local climate and its agricultural impacts.
2.2.3. Climate and growing season condition data
Several systems and tools leveraging state of the art
earth observation data are readily available and
have been deployed in national agricultural moni-
toring programs. The GEOGLAM Crop Monitor for
Early Warning (CM4EW) is a relatively recent crop
monitoring tool with the focus of providing
a reliable and vetted crop assessment for countries
exposed to food insecurity (Becker-Reshef et al.
2020). CM4EW provides monthly crop condition
assessments based on multi-source consensus data.
The monthly reports are prepared at the end of the
month, starting ten days before the publication date
to ensure timely information. Partner organizations
submit crop condition data which is then compli-
mented with agrometeorological earth observation
data at the sub-national level (Becker-Reshef et al.
2020). In addition to agricultural assessments, regio-
nal and global climatic conditions which are likely to
aect the growing season and crop yield are pro-
jected and outlined. A total of 88 monthly crop
condition reports and 6 special reports for the
study area are available for the 2016–2023 period
and were used in this study.
We used ERA5-Land reanalysis dataset provided by
the European Centre for Medium-Range Weather
Forecasts (ECMWF) for calculation of precipitation
and temperature in the research area (2019). Lastly,
the multivariate ENSO index was included to highlight
the evolution of El Niño and La Niña events which
Table 1. In-situ data statistics for push-pull and maize monocrop fields. Average
length, width and area of each crop type are shown, together with the minimum
and maximum dimensions within each crop type.
Cropping type Length, m Width, m Area, m2
Push-Pull 25.5 (10–57) 13.9 (10–25) 363.9 (100–918)
Maize monocrop 21.3 (10–42) 13.7 (10–24) 295.6 (120–816)
Table 2. A summary of the collected datasets used in this study. The satellite data consists of sentinel-2, Landsat 7, Landsat 8 and
Landsat 9 sensor data; the ground-truth vector data of the monocrop and push-pull fields were collected. Lastly, monthly climate
assessment reports from GEOGLAM crop monitor spanning 2016–2023 period were used.
Data Product Name Resolution Spatial-Temporal Source/references
Satellite data Sentinel-2 A/B 10m 5–6 days www.corpenicus.eu.
Landsat 7 15m 16 days www.usgs.gov.
Landsat 8 15m 16 days www.usgs.gov.
Landsat 9 15m 16 days www.usgs.gov.
Vector data Monocrop and Push-Pull ground-truth data field-level single time step fieldwork
Climate data CM4EW growing season condition assessment district-level monthly https://cropmonitor.org/.
Multivariate ENSO index no spatial resolution monthly (NCAR 2022)
ERA5-Land 9km daily [C3S (2019)]
GISCIENCE & REMOTE SENSING 5
signicantly inuence the weather pattern in East
Africa (NCAR 2022).
2.3. Methods
The study follows a 4-step methodological workow:
(i) growing season condition estimation: it focuses on
synthesizing monthly regional climate and growing
season assessment reports into a single, consistent
growing season condition overview from 2016 to
2023; (ii) satellite data pre-processing: it is centered
around satellite data pre-processing for phenological
metrics extraction; (iii) phenological metrics extrac-
tion: it utilizes a dense harmonized NDVI timeseries
to retrieve the phenological metrics of the maize
monocrop and Push-Pull intercrop elds in the study
area; and, (iv) statistical analysis of the impacts of
climate conditions on crop phenology: the phenolo-
gical outputs of sections 2 and 3 are compared with
the reported growing season conditions from sec-
tion 1.
2.3.1. Growing season condition estimation
The reports gathered from the GEOGLAM Crop
Monitor for Early Warning database were synthesized
to get an accurate depiction of the crop conditions
during the timeframe of this study. A special focus
was given on the identication and timing of extreme
hot, dry, or wet conditions, desert locust, FAW, and
delayed onset of the season. Growing seasons which
contained these conditions during the majority of the
season were categorized having stress-induced grow-
ing conditions, while growing seasons with reportedly
good conditions were categorized as having favour-
able growing conditions. The timing and cause of
stress-induced growing conditions were cross-
referenced with additional scientic publications
where available.
2.3.2. Satellite data pre-processing
The pre-processing of satellite data followed the
methodological workow presented in Liepa et al.
(2024). Several pre-processing steps were performed
to obtain a dense cloud-free and cloud shadow-free
time series data set spanning the years 2016 to 2023.
The image collections of Landsat 7 Enhanced
Thematic Mapper Plus (ETM+), Landsat 8 and 9
Operational Land Imager (OLI), and Sentinel-2
Multispectral Instrument (MSI) were ltered for our
study area and the study duration using general spa-
tial and temporal ltering methods in Google Earth
Engine (GEE) (Gorelick et al. 2017). Top-of-
Atmosphere (TOA) data collections were selected,
and the Sensor Invariant Atmospheric Correction
(SIAC) was applied (Yin, Lewis, and Gómez-Dans
2022). Utilizing the same atmospheric correction
method on imagery from dierent sensors allowed
us to minimize the discrepancies in atmospheric
eects exerted on the satellite imagery. Masking
cloud and cloud shadow was performed subse-
quently. Quality Assessment (QA) bands generated
from the CFMASK algorithm (Qiu, Zhu, and He 2019;
Zhu, Wang, and Woodcock 2015) were used to
remove pixels containing cloud contamination in the
Landsat sensors. Cloudy pixels in Sentinel-2 were
masked using the cloud probability band with the
probability value set to less than 25%. Pre-
processing was concluded by addressing the dierent
solar and view angles associated with satellite sen-
sors, by applying a Bi-directional Reectance
Distribution Function (BRDF) correction (Claverie
et al. 2018; Roy et al. 2016, 2017). This allowed us to
adjust the viewing and illumination angles of the
satellite imagery. The BRDF correction was executed
in the GEE cloud computing environment (Nguyen
et al. 2020; Poortinga et al. 2019).
2.3.3. Phenological metrics extraction
The dense NDVI time series was generated by harmo-
nizing Sentinel-2 A/B, Landsat 7, Landsat 8 and Landsat
9 satellite imagery (Liepa et al. 2024). A harmonic curve
was tted on the NDVI time series to smoothen the
observed data and cubic interpolation was applied for
gap lling purposes as proposed and implemented in
Google Earth Engine (GEE) by Descals et al. (2020).
A thresholding method was used to estimate the key
growing season parameters; start of season (SoS), end
of season (EoS) and duration of the season (DoS)
(Descals et al. 2021; Standfuß et al. 2022; Vrieling,
Leeuw, and Said 2013). The NDVI ratio was used for
setting the threshold value (White, Thornton, and
Running 1997). The ratio was calculated using the
absolute maximum and minimum values of the NDVI
during each growing season. The start of season (SoS)
and end of season (EoS) are denoted as the rst and
last days when the NDVI time series of a pixel exceeds
the local 50%-threshold (White et al. 2009). The dura-
tion of season is calculated by subtracting the date of
6A. LIEPA ET AL.
the EoS by the date of the SoS. In addition to the
before mentioned phenological metrics, the maximum
vegetation was captured by extracting the maximum
NDVI value during each growing season. The pheno-
logical extraction was applied on a pixel-by-pixel basis
and aggregated for each agricultural eld.
2.3.4. Statistical analysis of the impacts of climate
conditions on crop phenology
We investigated the signicance of the impacts of
climate conditions on crop phenology by applying
the Welch Two Sample t-test (Welch 1947). This
method allowed us to determine whether the crop
phonologies derived for the two agricultural systems
have signicant dierences. This test maintains nom-
inal type 1 error rates for groups with unequal sample
sizes and should thus produce a more robust statis-
tical analysis. For the t-test, we rejected the null
hypothesis of no signicant dierence at an error
probability of 0.1.
3. Results
3.1. Growing season conditions
A graphical summary of the crop conditions from
2016 to 2023 is shown in Figure 2. Southwestern
Kenya where the study region is situated experi-
ences more favorable agricultural conditions in con-
trast to other parts of Kenya. The proximity to Lake
Victoria makes the area more resilient to prolonged
dry spells and extensive heat episodes.
Nevertheless, this area is not immune to adverse
climate and periodically experiences poor, climate-
driven agricultural periods. The short growing sea-
son in 2016 was aected by delayed onset of rainfall
leading to dry conditions in October and November
(GEOGLAM Crop Monitor 2017a). Additionally, rst
instances of non-native fall armyworm in the
research area were reported (GEOGLAM Crop
Monitor 2017b). Delayed rainfall conditions were
documented in the subsequent short growing sea-
son of 2017 as well (GEOGLAM Crop Monitor 2017c).
Following mostly favorable short and long growing
seasons in 2018, the substantial rainfall decits
returned in 2019. February and March of 2019 regis-
tered up to 75% below-average cumulative rainfall
which drastically delayed crop planting in the area
(GEOGLAM Crop Monitor 2019). These signicantly
drier and hotter conditions were driven primarily by
the weak El Niño-Southern Oscillation (ENSO) con-
ditions. A year later, the Indian Ocean Dipole (IOD)
reversal enhanced rainfall in East Africa, causing
ooding in several counties in southwestern Kenya
(GEOGLAM Crop Monitor 2020). Above-average
rainfall caused an increase in vegetation even in
areas with previously sparse vegetation providing
favorable conditions for desert locust breeding
(Wang et al. 2021). Ultimately, this led to the worst
Desert Locust outbreak in East Africa in 25 years.
The most recent challenge came at the end of
the year 2021 when delayed rainfall and below-
average cumulative precipitation resulted in dimin-
ished yields in the 2021 short rainy season. The unfa-
vorable conditions persisted during the subsequent
long growing season at the beginning of 2022, which
suered greatly from unpredictable rains and exten-
sive dry spells (GEOGLAM Crop Monitor 2021). To
make matters worse, an African armyworm invasion
in late April 2022 caused extensive damage to crop
elds and added additional stress to an already vul-
nerable agricultural region (GEOGLAM Crop Monitor
2022).
3.2. Phenological metrics of the in-situ elds
The growing season durations of maize monocrop
and Push-Pull classes through the period of the
study are shown in Figure 3. Maize monocrop and
Push-Pull have very similar growing season duration
averages with a few exceptions being the 2018 long
growing season, the 2018 short growing season, the
2020 short growing season and the 2021 short grow-
ing season. Here, the growing seasons are longer in
the Push-Pull elds by about a week. The growing
seasons appear to be slightly longer for both agricul-
tural systems during the weak positive ENSO condi-
tion between mid-2018 and 2020 (Figure 2). Mean
and standard deviation values of growing season
duration and maximum NDVI are shown in
Appendix C.
Push-Pull elds recorded the highest maximum
NDVI averages on 11 of the total 15seasons
(Figure 4). During the long growing seasons of 2016
and 2017 as well as during the short growing season
of 2017 and 2018 the monocrop class showed the
highest NDVI averages. However, the biggest dier-
ence between monocrop and Push-Pull intercrop in
GISCIENCE & REMOTE SENSING 7
Figure 2. Graphical depiction of the growing season conditions between 2016 and mid-2023 for the study area based on monthly
GEOGLAM Crop Monitor assessments. Multivariate ENSO index (NCAR 2022) is depicted with negative (red) and positive (green)
phases indicating El Niño and La Niña events respectively. Monthly temperature and precipitation data acquired from ERA5-land
dataset (2019) together with the long-term mean in red.
8A. LIEPA ET AL.
the context of maximum NDVI does not appear dur-
ing these seasons. The biggest discrepancy is
observed during the short growing seasons of 2016
and 2022 where Push-Pull intercrop showed higher
maximum NDVI values by 0.5 and 0.4, respectively.
Interestingly, the biggest discrepancies occur during
the short growing seasons which are known to experi-
ence less predictive and more sporadic rains.
The growing seasons since 2022 have shown con-
sistently low maximum NDVI values for all classes. The
Figure 3. Growing season duration violin plots for maize monocrop and push-pull intercrop classes derived using the harmonized
sentinel-2 and Landsat product. Maize monocrop class is depicted in green and the push-pull intercrop is visible in blue. Black dot
indicates the mean value of each class. Seasons with stress-induced growing conditions are depicted by yellow boxes and are based
on Figure 2.
Figure 4. Growing season maximum NDVI violin plots for each growing season derived using the harmonized sentinel-2 and Landsat
product. Maize monocrop class is depicted in green and the push-pull intercrop is visible in blue. Black dot indicates the mean value of
each class. Seasons with stress-induced growing conditions are depicted by yellow boxes and are based on Figure 2.
GISCIENCE & REMOTE SENSING 9
same is partially reected in the duration of the sea-
sons as both maize monocrop and Push-Pull have
recorded short growing season durations since 2022
(Figure 3).
3.3. Eects of climatic forcings on push-pull and
monocrop elds
The growing conditions were unfavorable in two of
the four seasons with the most dierence in seasonal
duration (Figure 5). This was due to the desert locust
outbreak (2020 short season) in combination with the
fall armyworm and delayed rainfall, and below-
average precipitation (2021 short season). Here, Push-
Pull elds recorded longer growing season duration
than its monocrop counterparts. During the growing
seasons with favorable environment, Push-Pull and
maize monocrop elds had very even growing season
lengths with one exception being 2018 long where
Push-Pull agricultural cycle lasted 97 days in contrast
to 88 days of maize monocrop.
In four periods where monocrop elds had longer
growing seasons, two of the periods had unfavorable
climatic conditions caused by delayed rainfall (2017
short season) and drought (2019 long season). The
average day dierence recorded was 1 and 3 days,
respectively. Favorable growing conditions were
registered during the season with the largest
negative day dierence between Push-Pull and
Monocrop (2021 long season). This is also the only
season with statistical analysis indicating a signicant
dierence in duration of season with a p-value of
0.034. The p-values of the Welch Two Sample t-test
for all growing seasons and phenological metrics are
presented in the Appendix D.
Highest maximum NDVI dierences were recorded
during the short growing seasons coinciding with
unfavorable growing conditions (Figure 6). The 2016
short and 2021 short growing seasons featured unfa-
vorable growing conditions caused by fall armyworm
outbreaks, the pest against which the Push-Pull agri-
cultural concept was developed to combat. The
p-values of 0.034 and 0.013, respectively, support
the signicant dierence in maximum NDVI dier-
ence for the short growing seasons of 2016 and
2021. The short growing season of 2017 featured
delayed rainfall. With the p-value of 0.367, the
maxNDVI dierence for this season is not supported
by the statistical analysis. Among the seasons with the
highest maxNDVI dierences, the short season of
2022 (p-value of 0.005) appears to be the only one
occurring during favorable growing conditions.
Both agriculture types show some degree of eect
by climatic forcings in terms of start and end of grow-
ing season (Figure 7). During the short rain seasons of
2016, 2017, 2020 and 2021 which featured stress-
induced growing conditions, the end of season
occurred considerably earlier in both agricultural sys-
tems. The same is not apparent during the long rainy
seasons where the seasons with stress-induced con-
ditions (years 2019, 2020 and 2022) show later onset
and end of the season. Long growing seasons of 2019
and 2022 recorded substantially late onsets for both
cropping systems with seasonal onsets starting at the
end of April to beginning of May. These recordings
correspond with documented considerable rain de-
cits and delayed onset of seasonal rains.
Inter-class comparison shows that the start of sea-
son for maize monocrop and Push-Pull show signi-
cant dierences during the 2017 and 2020 short rainy
seasons, and the long rainy seasons of 2019 and 2020.
These seasons feature stress-induced growing condi-
tions. Weather induced unfavorable growing condi-
tions produced signicant seasonal onset dierence
between maize monocrop and Push-Pull agriculture.
The dierence in the end of seasons between the two
agricultural classes is signicant during the 2016 and
2022 short rainy seasons, and the long rainy seasons
of 2016 and 2023. Of these four seasons, only the
short rainy seasons possessed stress-induced growing
conditions. This indicates that the end of the growing
season is inuenced by fall armyworm and delayed
rainfall.
Stable conditions with little to no dierence
between start and end of growing seasons are regis-
tered in favorable growing settings as well. Push-Pull
elds in general show a later end of season, however
the dierence is not signicant.
4. Discussion
4.1. Climatic condition assessment
The synthesis of the climatological assessment
reports from GEOGLAM provided an overview of
weather conditions in the research area. This allowed
us to determine whether the growing conditions
were favorable or stress-induced in terms of rain
10 A. LIEPA ET AL.
Figure 5. Difference in seasonal durations between push-pull and monocrop classes. The positive values in green indicate seasons
during which the push-pull crop type observed longer seasonal duration average. The negative values in red indicate the seasons
where push-pull showed shorter seasonal duration average.
Figure 6. Difference in seasonal maximum NDVI between push-pull and monocrop classes. The positive values in green indicate
seasons where push-pull class observed on average higher maxNDVI values. Higher NDVI values have been observed during all
growing seasons except 2017 long season where NDVI values were equal.
GISCIENCE & REMOTE SENSING 11
timing and intensity as well as temperature and biotic
stresses. The literature review revealed that 7 out of
the total 15 growing seasons had stress-induced
growing conditions which may have exerted unfavor-
able conditions on the crops in the research area. The
outputs of GEOGLAM reports in the area are consis-
tent with other assessments. The rainfall anomalies of
October to December 2016 and rst outbreaks of fall
armyworm were highlighted by Uhe et al. (2018) and
Groote et al. (2020) respectively. Rain decits of the
2019 and the 2022 long growing seasons and delayed
onset of the 2017 short season were also reported
(Funk et al. 2022; Han et al. 2022; Harrison and Way-
Henthorne 2019). Excessive rainfall which caused
ooding and subsequent crop failures of 2020 were
mentioned in a large meteorological study by
Wainwright et al. (2021). A desert locust outbreak
exerted additional stress to crops in the end of 2020
(Kimathi et al. 2020; Mullié et al. 2023). Lastly, the
most recent event of 2021/2022 resulting in adverse
eects on agriculture inicted by reappearance of the
fall armyworm, delay in rainfall and below-average
rain totals were documented in several studies (Funk
et al. 2022; Kansiime, Rwomushana, and Mugambi
2023; Mutyambai et al. 2022). Overall, these studies
conrm the climate conditions assumed here on the
basis of GEOGLAM reports.
4.2. Seasonal phenological metrics estimation
Using the harmonized dataset of Landsat 7, 8, 9, and
Sentinel-2 together with the NDVI-based thresholding
method allowed us to extract phenological metrics of
maize monocrop and Push-Pull intercrop elds. This
marks the rst attempt in comparing traditional agri-
culture with agricultural systems featuring applied
Figure 7. Start of season (SoS) in brighter colors and end of season (EoS) in darker colors for each class during the study period derived
using the harmonized sentinel-2 and Landsat product. Maize monocrop class is depicted in green, push-pull intercrop is visible in blue.
Black dot indicates the mean value of each class and seasons with stress-induced growing conditions are depicted by yellow boxes
and are based on Figure 2.
12 A. LIEPA ET AL.
ecological management strategies based on earth
observation data. For reference on the seasonal tim-
ing, we planned the ground truth data gathering
campaign during the early stage of 2023 long grow-
ing season. Field observations gave us a clear timing
on the start of the growing season. Farmer interviews
added further knowledge on the growing status and
anticipated harvest. The monthly assessments from
GEOGLAM CM4EW were additional sources used for
the seasonal timing reference. Based on these
sources, our phenology extraction method could be
ne-tuned.
The phenology extraction method allowed us to
retrieve the start, peak, duration and end of each
growing season. The descriptive ability of the pheno-
logical metrics is not uniform. In terms of capturing
the changes in phenology driven by climatic forcings
between the monocrop and Push-Pull agricultural
classes, maximum NDVI provided most descriptive
value (Liepa et al. 2024). This is not surprising as
maximum NDVI targets the detection of vegetation
productivity in turn shedding light on the health of
plants in a eld. In contrast, the duration of season
showed minimal dierence between monocrop and
Push-Pull agriculture. This can be attributed to the
farming practices, as the farmers tend to clear the
elds toward the end of the season regardless of the
crop yield. Even in crop failure during the early stages
of the season, the plants remain in the elds as
removal of them can lead to soil degradation and
less moisture retention. This impacts the descriptive
ability of the end of season metric as well (Liepa et al.
2024). Little change was detected between classes in
terms of start of the season since both agricultural
types exhibited similar onsets of the season through-
out the study period.
The harmonization performance was analyzed on
10-m resolution by resampling the Landsat imagery.
By following the pre-processing steps outlined in Liepa
et al. (2024), the rescaling prevented loss of data qual-
ity. The combined use of Landsat and Sentinel-2 pro-
vided a cloud-free image for each pixel every three to
four days. The use of the Sentinel-2 constellation alone
would provide a revisit time of ve days prior to cloud
and cloud shadow masking. Thus, inclusion of the
Landsat sensors greatly increases the temporal cover-
age of the target areas.
A noteworthy constraint of the methodology is the
relatively small sample size. While the Push-Pull
intercrop is now a widely employed cropping practice
in Kenya, there are very few elds that have been
cultivated throughout the duration of the study. This
renders the comparison of the climatic impacts on
Push-Pull intercrop and conventional maize mono-
crop. The sample size also has an eect on the results
of statistical signicance. A small sample size
decreases the power of the statistical tests by increas-
ing the risk of Type II errors, making the true eects
dicult to detect. This might explain the misalign-
ment between observed notable dierences in phe-
nological metrics and the statistical signicance
outlined in section 3.3
4.3. Relationship between the phenological metrics
and seasonal climatic forcings
Dierences between maize monocrop and Push-Pull
intercrop elds were at their peak during periods of
biotic stresses in the area. These were caused by
desert locust (2020 short season) and fall armyworm
(2016 short, 2021 short and 2022 long seasons). The
FAW outbreaks were accompanied by delayed and
below-average rainfall in all three seasons.
Interestingly, the rst outbreak of the fall armyworm
recorded in the 2016 short season showed similar
dierence in maximum NDVI values, but the duration
of the season dierence was less profound. This
shows that the response of these two agricultural
systems to biotic stresses is not always the same
from a remote sensing perspective. Such behavior
can have several explanations, including transient on-
farm farmer mitigative interventions and model per-
formance. As the 2016 short season saw the rst
recorded outbreak of FAW in the area, farmer inter-
vention via pesticide or fertilizer use could have
deviated from later outbreaks. From the modeling
perspective, less EO data further back in time might
have impacted the harmonization eorts leading to
diminished phenology extraction accuracy.
The biggest discrepancies between the two agri-
cultural systems occurred mainly during the short
growing season between the October and
December months. Unlike the rains during the long
growing season, rainfall in the period of October to
December is more unpredictable which can cause
issues with crop planting for farmers. Having Push-
Pull outperforming maize monocrop during short
growing seasons may indicate a higher adaptive
GISCIENCE & REMOTE SENSING 13
capability of Push-Pull intercropping. This is further
supported by a study based on household surveys in
the research area by Ndayisaba et al. (2023). The
research found that under rainfall-decit periods
Push-Pull agriculture yielded more maize than its non-
push-pull counterpart. Higher climate resilience was
also recorded in Ethiopia by Gugissa, Abro, and Tefera
(2022) highlighting that the ndings are not site-
specic.
4.4. Future research and outlooks
The current study emphasizes changes in precipita-
tion and temperature as direct eects of climate
change on agriculture. In the context of the perfor-
mance of Push-Pull and non-Push-Pull agriculture
under future climate, it is important to note additional
inuences such as below-ground processes
(Rosenzweig et al. 2001; Ziska and McConnell 2016),
soil nutrients and organic matter (Chen et al. 2020; Lu
et al. 2013) and soil microbial biomass (Classen et al.
2015). Our research demonstrates the added value of
earth observation tools in agricultural monitoring of
small-scale agriculture in East Africa. Because of this,
adaptation strategies should consider the combined
scientic outputs spanning several research disci-
plines including earth observation. Such future ecolo-
gical management strategies should also seek to
meet various farmer needs and priorities including
dietary diversity and address other constraints to
ensure value addition and scaling (Chidawanyika
et al. 2023). For the Push-Pull, the innovative
approach has evolved over the years from the con-
ventional to climate smart and later the 3rd genera-
tion Push-Pull technology to address both biotic and
abiotic constraints (Cheruiyot et al. 2021). This has
ensured wider adoption of the Push-Pull, even to
much more arid regions based on context-specic
companion cropping. More recently, the Push-Pull
has been further intensied by integration with vege-
tables and edible legumes (Chidawanyika et al. 2023).
This approach not only helps in building resilience,
but also support scaling through bundled benets
that address various needs.
The study also shows that an area-wide EO-based
monitoring system has the potential to systematically
monitor and evaluate the acute situation of agricul-
tural production. To fully understand the capabilities
of such methodologies, future research should focus
on transferability of the method to other regions
featuring dierent pest and climate proles.
5. Conclusion
This study marks the rst attempt in comparing tradi-
tional maize cropping with agricultural systems fea-
turing applied ecological management strategies
based on earth observation and in situ data. Our
study concludes with the following ndings:
(i) Climate assessment of the area based on avail-
able climatological resources shows that 7 out
of 15 growing seasons experienced unfavorable
growing conditions between the years 2016 and
2023. These were caused by climatic factors,
pest outbreaks or a combination of both.
(ii) Our research introduces a temporal element in
agricultural monitoring by harmonizing Landsat
7, 8, 9 with Sentinel-2 spanning the duration of
the study. In doing so, phenological metrics of
the growing seasons are extracted based on
a threshold method using the NDVI as
a vegetation proxy. The comparison reveals
that inter-class dierences exist and are detect-
able using earth observation tools.
(iii) The impact of regional climatic forcings on the
two agricultural systems was most severe dur-
ing periods of biotic stresses caused by desert
locust outbreak and fall armyworm. Higher
values in both duration of the season and max-
imum NDVI indicated a higher resilience by
Push-Pull agriculture in times of unfavorable
agricultural conditions. Start and end of grow-
ing seasons for both cropping systems show
similar responses to weather. Short rainy sea-
sons with unfavorable growing conditions
resulted in an earlier end of season. Long grow-
ing seasons with unfavorable conditions
brought about delayed growing season start
and ends.
Our research demonstrates the potential of earth
observation tools in agricultural monitoring.
Unpredictability in the future climate and forecasted
increase in abrupt temperature and rainfall variations
will inuence the planting patterns and growing con-
ditions of crops. Future management strategies
should consider the combined scientic outputs
14 A. LIEPA ET AL.
spanning several research disciplines including earth
observation to enhance the adaptive capacity and
resilience of cropping systems.
Acknowledgments
The authors express gratitude to the U.S. Geological Survey
(USGS) Earth Resources Observation and Science (EROS) Center
for Landsat 7, 8 and 9 imagery; the Copernicus Sentinel-2
mission for the Sentinel-2 data; the GEOGLAM Crop Monitor
for CM4EW monthly climate assessment reports; NCAR for
Multivariate ENSO index; C3S for ERA5-Land precipitation and
temperature data; Google Earth Engine platform for its cloud-
computing capabilities; and the European Commission's
Horizon 2020 research and innovation program for funding
via the project UPSCALE (https://upscale-h2020.eu/). Lastly,
the authors would like to thank the anonymous reviewers for
their constructive feedback and insightful recommendations.
Disclosure statement
No potential conict of interest was reported by the author(s).
Funding
This publication is supported by the Open Access Publication
Fund of the University of Wuerzburg. The research is a part of
the project UPSCALE funded by the European Commission’s
Horizon 2020 research and innovation program under grant
agreement [No 861998].
ORCID
Adomas Liepa http://orcid.org/0009-0004-0136-8584
Marcell K. Peters http://orcid.org/0000-0002-1262-0827
CRediT authorship contribution statement
Adomas Liepa: Conceptualization, Data curation, Formal analy-
sis, Investigation, Methodology, Software, Visualization, Writing –
original draft, Writing – review & editing, Resources. Michael
Thiel: Conceptualization, Funding acquisition, Project administra-
tion, Resources, Supervision, Writing – review & editing. Hannes
Taubenböck: Resources, Writing – review & editing. Doris Klein:
Resources, Writing – review & editing. Ingolf Stean-Dewenter:
Resources, Funding acquisition, Project administration. Sarah
Schönbrodt-Stitt: Resources, Writing – review & editing. Insa
Otte: Resources, Writing – review & editing. Tobias Landmann:
Resources, Writing – review & editing. Michael Ochieng
Obondo: Resources. Zeyaur R. Khan: Funding acquisition.
Frank Chidawanyika: Funding acquisition, Project administra-
tion, Resources. Emily A. Martin: Funding acquisition, Project
administration, Resources. Tobias Ullmann: Conceptualization,
Resources, Supervision, Writing – review & editing.
Data availability statement
Data will be made available on request.
References
Adams, R. M., B. H. Hurd, S. Lenhart, and N. Leary. 1998. “Eects
of Global Climate Change on World Agriculture: An
Interpretive Review.” Climate Research 11 (1): 19–30.
https://doi.org/10.3354/cr011019 .
Allen, M. 2003. “Liability for Climate Change.” Nature
421 (6926): 891–892. https://doi.org/10.1038/421891a .
Aydinalp, C., and M. S. Cresser. 2008. “The Eects of Global
Climate Change on Agriculture.” American-Eurasian Journal
of Agriculture & Environmental Science.3 (5) :672–676 https://
www.researchgate.net/prole/malcolm-cresser/publication/
238091112_the_eects_of_global_climate_change_on_agri
culture/links/02e7e532ca03f62d72000000/the-eects-of-
global-climate-change-on-agriculture.pdf .
Beck, H. E., N. E. Zimmermann, T. R. McVicar, N. Vergopolan,
A. Berg, and E. F. Wood. 2018. “Present and Future Köppen-
Geiger Climate Classication Maps at 1-Km Resolution.”
Scientic Data 5 (1): 180214. https://doi.org/10.1038/sdata.
2018.214 .
Becker-Reshef, I., C. Justice, B. Barker, M. Humber, F. Rembold,
R. Bonifacio, M. Zappacosta, et al. 2020. “Strengthening
Agricultural Decisions in Countries at Risk of Food
Insecurity: The GEOGLAM Crop Monitor for Early Warning.”
Remote Sensing of Environment 237:111553. https://doi.org/
10.1016/j.rse.2019.111553 .
Buleti, S. I., S. Kuyah, A. Olagoke, M. Gichua, S. Were,
F. Chidawanyika, and E. A. Martin. 2023. “Farmers’
Perceived Pathways for Further Intensication of Push-Pull
Systems in Western Kenya.” Frontiers in Sustainable Food
Systems. 7, Article 1191038. 7. https://doi.org/10.3389/fsufs.
2023.1191038 .
C3S. 2019. ERA5-Land Monthly Averaged Data from 1950 to
Present. Accessed March 25, 2024. https://doi.org/10.24381/
cds.68d2bb30
Challinor, A. J., J. Watson, D. B. Lobell, S. M. Howden,
D. R. Smith, and N. Chhetri. 2014. “A Meta-Analysis of Crop
Yield Under Climate Change and Adaptation.” Nature
Climate Change 4 (4): 287–291. https://doi.org/10.1038/ncli
mate2153 .
Chen, J., L. Elsgaard, van Groenigen, J. E. Olesen, K. Jan,
J. E. Olesen, Z. Liang, et al. 2020. “Soil Carbon Loss with
Warming: New Evidence from Carbon-Degrading
Enzymes.” Global Change Biology 26 (4): 1944–1952.
https://doi.org/10.1111/gcb.14986 .
Cheruiyot, D., F. Chidawanyika, C. A. O. Midega, J. O. Pittchar,
J. A. Pickett, and Z. R. Khan. 2021. “Field Evaluation of a New
Third Generation Push-Pull Technology for Control of Striga
GISCIENCE & REMOTE SENSING 15
Weed, Stemborers, and Fall Armyworm in Western Kenya.”
Experimental Agriculture 57 (5–6): 301–315. https://doi.org/
10.1017/S0014479721000260 .
Chidawanyika, F., B. Muriithi, S. Niassy, F. O. Ouya, J. O. Pittchar,
M. Kassie, and Z. R. Khan. 2023. “Sustainable Intensication
of Vegetable Production Using the Cereal ‘Push-Pull tech-
nology’: Benets and One Health Implications.”
Environmental Sustainability 6 (1): 25–34. https://doi.org/10.
1007/s42398-023-00260-1 .
Classen, A. T., M. K. Sundqvist, J. A. Henning, G. S. Newman,
J. A. M. Moore, M. A. Cregger, L. C. Moorhead, and
C. M. Patterson. 2015. “Direct and Indirect Eects of
Climate Change on Soil Microbial and Soil Microbial‐Plant
Interactions: What Lies Ahead?” Ecosphere 6 (8): 1–21.
https://doi.org/10.1890/ES15-00217.1 .
Claverie, M., J. Ju, J. G. Masek, J. L. Dungan, E. F. Vermote, J.-
C. Roger, S. V. Skakun, and C. Justice. 2018. “The Harmonized
Landsat and Sentinel-2 Surface Reectance Data Set.”
Remote Sensing of Environment 219:145–161. https://doi.
org/10.1016/j.rse.2018.09.002 .
Cressman, K. 2013. “Role of Remote Sensing in Desert Locust
Early Warning.” Journal of Applied Remote Sensing 7 (1):
75098. https://doi.org/10.1117/1.JRS.7.075098 .
David, O. G., A. S. Ayangbenro, J. J. O. Odhiambo, and
O. O. Babalola. 2022. “Striga Hermonthica: A Highly
Destructive Pathogen in Maize Production.” Environmental
Challenges 8:100590. https://doi.org/10.1016/j.envc.2022.
100590 .
Descals, A., A. Verger, G. Yin, and J. Penuelas. 2021.
“A Threshold Method for Robust and Fast Estimation of
Land-Surface Phenology Using Google Earth Engine.” IEEE
Journal of Selected Topics in Applied Earth Observations &
Remote Sensing 14:601–606. https://doi.org/10.1109/
JSTARS.2020.3039554 .
Descals, A., A. Verger, G. Yin, and J. Peñuelas. 2020. “Improved
Estimates of Arctic Land Surface Phenology Using Sentinel-2
Time Series.” Remote Sensing 12 (22): 3738. https://doi.org/
10.3390/rs12223738 .
Deschenes, O., and M. Greenstone. 2004 The Economic Impacts
of Climate Change: Evidence from Agricultural Prots and
Random. American Economic Review 97 (1): 354–385 doi:10.
1257/aer.97.1.354
Ekesa, B., J. Rao, J.-J. Cadilhon, P. Ayebare, B. Bashaasha,
C. Muyanja. 2015. A Situational Analysis of Agricultural
Production and Marketing, and Natural Resource
Management Systems in the Central Region of Uganda.
doi:10.13140/RG.2.1.3559.8962 .
Funk, C., G. Galu, D. Korecha, L. Harrison, W. Anderson, A. Hoell.
2022. “Forecast Update – East Africa Likely to Experience Six
Droughts in a Row. University of California, Santa Barbara’s
Climate Hazards Center (CHC). Accessed January 15, 2024.
https://blog.chc.ucsb.edu/?p=1240 .
Gao, F., and X. Zhang. 2021. “Mapping Crop Phenology in Near
Real-Time Using Satellite Remote Sensing: Challenges and
Opportunities.” Journal of Remote Sensing 2021, Article
2021/8379391. https://doi.org/10.34133/2021/8379391 .
Gebrechorkos, S. H., S. Hülsmann, and C. Bernhofer. 2019.
“Long-Term Trends in Rainfall and Temperature Using
High-Resolution Climate Datasets in East Africa.” Scientic
Reports 9 (1): 11376. https://doi.org/10.1038/s41598-019-
47933-8 .
GEOGLAM. 2019. “Special Report: Kenya Long Rains Rapid Crop
Assessment. GEOGLAM Crop Monitor. Accessed October 26,
2023 https://cropmonitor.org/index.php/2019/09/24/spe
cial-update-kenya/ .
GEOGLAM Crop Monitor. 2017a. Early Warning Crop Monitor
No. 13. GEOGLAM (13). Accessed March 11, 2023. https://
cropmonitor.org/documents/EWCM/reports/EarlyWarning_
CropMonitor_201702.pdf .
GEOGLAM Crop Monitor. 2017b. “Early Warning Crop Monitor
No. 16.” GEOGLAM (16). Accessed March 11, 2023. https://
cropmonitor.org/documents/EWCM/reports/EarlyWarning_
CropMonitor_201705.pdf .
GEOGLAM Crop Monitor. 2017c. “Early Warning Crop Monitor
No. 23.” GEOGLAM (23). Accessed March 11, 2023. https://
cropmonitor.org/documents/EWCM/reports/EarlyWarning_
CropMonitor_201712.pdf .
GEOGLAM Crop Monitor. 2019. “Early Warning Crop Monitor
No. 37.” GEOGLAM (37). Accessed March 11, 2023. https://
cropmonitor.org/documents/EWCM/reports/EarlyWarning_
CropMonitor_201904.pdf .
GEOGLAM Crop Monitor. 2020. Early Warning Crop Monitor No.
46. GEOGLAM (46). Accessed March 11, 2023. https://crop
monitor.org/documents/EWCM/reports/EarlyWarning_
CropMonitor_202002.pdf .
GEOGLAM Crop Monitor. 2021. “Earl Warning Crop Monitor No.
66.” GEOGLAM (66). Accessed March 11, 2023. https://crop
monitor.org/documents/EWCM/reports/EarlyWarning_
CropMonitor_202111.pdf .
GEOGLAM Crop Monitor. 2022. “Early Warning Crop Monitor
No. 71.” GEOGLAM (71). Accessed March 11, 2023. https://
cropmonitor.org/documents/EWCM/reports/EarlyWarning_
CropMonitor_202205.pdf .
Gorelick, N., M. Hancher, M. Dixon, S. Ilyushchenko, D. Thau,
and R. Moore. 2017. “Google Earth Engine: Planetary-Scale
Geospatial Analysis for Everyone.” Remote Sensing of
Environment 202:18–27. https://doi.org/10.1016/j.rse.2017.
06.031 .
Groote, H. D., S. C. Kimenju, B. Munyua, S. Palmas, M. Kassie,
and A. Bruce. 2020. “Spread and Impact of Fall Armyworm
(Spodoptera Frugiperda J.E. Smith) in Maize Production
Areas of Kenya.” Agriculture, Ecosystems & Environment
292:106804. https://doi.org/10.1016/j.agee.2019.106804 .
Gugissa, D. A., Z. Abro, and T. Tefera. 2022. “Achieving
a Climate-Change Resilient Farming System Through Push–
Pull Technology: Evidence from Maize Farming Systems in
Ethiopia.” Sustainability 14 (5): 2648. https://doi.org/10.3390/
su14052648 .
Halford, N. G., T. Y. Curtis, Z. Chen, and J. Huang. 2015. “Eects
of Abiotic Stress and Crop Management on Cereal Grain
Composition: Implications for Food Quality and Safety.”
Journal of Experimental Botany 66 (5): 1145–1156. https://
doi.org/10.1093/jxb/eru473 .
16 A. LIEPA ET AL.
Han, X., Y. Li, W. Yu, and L. Feng. 2022. “Attribution of the
Extreme Drought in the Horn of Africa During Short-Rains
of 2016 and Long-Rains of 2017.” Water 14 (3): 409. https://
doi.org/10.3390/w14030409 .
Harrison, L., and J. Way-Henthorne. 2019. “Kenya’s March-
To-May Season: An Examination of April 2019
Predictions, May 2019 Outcomes, and the Implications
of a Poor Start to the Season.” University of California,
Santa Barbara’s Climate Hazards Center (CHC). Accessed
January 15, 2019. https://agrilinks.org/post/kenyas-march-
may-season-examination-april-2019-predictions-may-2019-
outcomes-and-implications .
Henrys, P. A., and S. G. Jarvis. 2019. “Integration of Ground
Survey and Remote Sensing Derived Data: Producing
Robust Indicators of Habitat Extent and Condition.” Ecology
and Evolution 9 (14): 8104–8112. https://doi.org/10.1002/
ece3.5376 .
Hooper, A. M., J. C. Cauleld, B. Hao, J. A. Pickett,
C. A. O. Midega, and Z. R. Khan. 2015. “Isolation and
Identication of Desmodium Root Exudates from Drought
Tolerant Species Used as Intercrops Against Striga
Hermonthica.” Phytochemistry 117:380–387. https://doi.org/
10.1016/j.phytochem.2015.06.026 .
Hulme, M., R. Doherty, T. Ngara, M. New, and D. Lister. 2001.
“African Climate Change: 1900-2100.” Climate Research
17:145–168. https://doi.org/10.3354/cr017145 .
Hütsch, B. W., D. Jahn, and S. Schubert. 2019. “Grain Yield of
Wheat (Triticum Aestivum L.) Under Long-Term Heat Stress
is Sink-Limited with Stronger Inhibition of Kernel Setting
Than Grain Filling.” Journal of Agronomy and Crop Science
205 (1): 22–32. https://doi.org/10.1111/jac.12298 .
IPCC. 2023. Climate Change 2023: Synthesis Report. Contribution
of Working Groups I, II and III to the Sixth Assessment Report of
the Intergovernmental Panel on Climate Change [Core Writing
Team, H. Lee and J. Romero (Eds.)]. Geneva, Switzerland: IPCC.
Jafari Jozani, H., M. Thiel, E. M. Abdel-Rahman, K. Richard,
T. Landmann, S. Subramanian, and M. Hahn. 2022.
“Investigation of Maize Lethal Necrosis (MLN) Severity and
Cropping Systems Mapping in Agro-Ecological Maize
Systems in Bomet, Kenya Utilizing RapidEye and Landsat-8
Imagery.” Geology, Ecology, and Landscapes 6 (2): 125–140.
https://doi.org/10.1080/24749508.2020.1761195 .
Kansiime, M. K., I. Rwomushana, and I. Mugambi. 2023. “Fall
Armyworm Invasion in Sub-Saharan Africa and Impacts on
Community Sustainability in the Wake of Coronavirus
Disease 2019: Reviewing the Evidence.” Current Opinion in
Environmental Sustainability 62:101279. https://doi.org/10.
1016/j.cosust.2023.101279 .
Kr, R., W. A. Overholt, Z. R. Khan, and A. Polaszek. 2002.
“Biology and Management of Economically Important
Lepidopteran Cereal Stem Borers in Africa.” Annual Review
of Entomology 47 (1): 701–731. https://doi.org/10.1146/
annurev.ento.47.091201.145254 .
Khan, Z., C. A. O. Midega, A. Hooper, and J. Pickett. 2016. “Push-
Pull: Chemical Ecology-Based Integrated Pest Management
Technology.” Journal of Chemical Ecology 42 (7): 689–697.
https://doi.org/10.1007/s10886-016-0730-y .
Khan, Z. R., A. Hassanali, W. Overholt, T. M. Khamis,
A. M. Hooper, J. A. Pickett, L. J. Wadhams, and
C. M. Woodcock. 2002. “Control of Witchweed Striga
Hermonthica by Intercropping with Desmodium Spp. and
the Mechanism Dened as Allelopathic.” Journal of Chemical
Ecology 28 (9): 1871–1885. https://doi.org/10.1023/
A:1020525521180 .
Khan, Z. R., J. A. Pickett, L. Wadhams, and F. Muyekho. 2001.
“Habitat Management Strategies for the Control of Cereal
Stemborers and Striga in Maize in Kenya.” International
Journal of Tropical Insect Science 21 (4): 375–380. https://
doi.org/10.1017/S1742758400008481 .
Kimathi, E., H. E. Z. Tonnang, S. Subramanian, K. Cressman,
E. M. Abdel-Rahman, M. Tesfayohannes, S. Niassy, et al.
2020. “Prediction of Breeding Regions for the Desert
Locust Schistocerca Gregaria in East Africa.” Scientic
Reports 10 (1): 11937. https://doi.org/10.1038/s41598-020-
68895-2 .
Liepa, A., M. Thiel, H. Taubenböck, I. Stean-Dewenter, I.-
O. Abu, S. Dhillon, Otte, et al. 2024. “Harmonized NDVI
Time-Series from Landsat and Sentinel-2 Reveal
Phenological Patterns of Diverse, Small-Scale Cropping
Systems in East Africa.” Remote Sensing Applications: Society
& Environment 35:101230. https://doi.org/10.1016/j.rsase.
2024.101230 .
Lu, M., X. Zhou, Q. Yang, H. Li, Y. Luo, C. Fang, J. Chen, X. Yang,
and B. Li. 2013. “Responses of Ecosystem Carbon Cycle to
Experimental Warming: A Meta-Analysis.” Ecology 94 (3):
726–738. https://doi.org/10.1890/12-0279.1 .
Main-Knorn, M., B. Pug, J. Louis, V. Debaecker, U. Müller-Wilm,
and F. Gascon. 2017. “Sen2Cor for Sentinel-2.” Proceedings of
the SPIE 10427 (10427). Accessed October, 2017. https://
www.spiedigitallibrary.org/conference-proceedings-of-spie
/10427/2278218/Sen2Cor-for-Sentinel-2/10.1117/12.
2278218.full .
Malhi, G. S., M. Kaur, and P. Kaushik. 2021. “Impact of Climate
Change on Agriculture and Its Mitigation Strategies: A
Review.” Sustainability 13 (3): 1318. https://doi.org/10.3390/
su13031318 .
Market Aliance. 2022. “Kenya Country Brief. Farm to Market
Alliance. Accessed October 27, 2023. https://ftma.org/wp-
content/uploads/2022/02/FtMA_Kenya_CountryBrief_Dec-
2022.pdf .
Midega, C. A. O., T. J. A. Bruce, J. A. Pickett, and Z. R. Khan. 2015.
“Ecological Management of Cereal Stemborers in African
Smallholder Agriculture Through Behavioural Manipulation.”
Ecological Entomology 40 (Suppl 1): 70–81. https://doi.org/10.
1111/een.12216 .
Mubenga-Tshitaka, J.-L., J. Dikgang, M. Mwamba, W. John, and
D. Gelo. 2023. “Climate Variability Impacts on Agricultural
Output in East Africa.” Cogent Economics and Finance 11 (1),
Article 2181281. https://doi.org/10.1080/23322039.2023.
2181281 .
Mullié, W. C., A. Prakash, A. Müller, and E. Lazutkaite. 2023.
“Insecticide Use Against Desert Locust in the Horn of Africa
2019–2021 Reveals a Pressing Need for Change.” Agronomy
13 (3): 819. https://doi.org/10.3390/agronomy13030819 .
GISCIENCE & REMOTE SENSING 17
Mumo, L., J. Yu, and K. Fang. 2018. “Assessing Impacts of
Seasonal Climate Variability on Maize Yield in Kenya.”
International Journal of Plant Production 12 (4): 297–307.
https://doi.org/10.1007/s42106-018-0027-x .
Mutyambai, D. M., S. Niassy, P.-A. Calatayud, and
S. Subramanian. 2022. “Agronomic Factors Inuencing Fall
Armyworm (Spodoptera frugiperda) Infestation and
Damage and Its Co-Occurrence with Stemborers in Maize
Cropping Systems in Kenya.” Insects 13 (3): 266. https://doi.
org/10.3390/insects13030266 .
Nakalembe, C., I. Becker-Reshef, R. Bonifacio, G. Hu,
M. L. Humber, C. J. Justice, J. Keniston, et al. 2021.
“A Review of Satellite-Based Global Agricultural Monitoring
Systems Available for Africa.” Global Food Security
29:100543. https://doi.org/10.1016/j.gfs.2021.100543 .
NCAR. 2022. “The Climate Data Guide: Multivariate ENSO Index.
Last Modied 2022-11-07.” National Center for Atmospheric
Research. Accessed May, 2024. https://climatedataguide.
ucar.edu/climate-data/multivariate-enso-index .
Ndayisaba, P. C., S. Kuyah, C. A. O. Midega, P. N. Mwangi, and
Z. R. Khan. 2023. “Push-Pull Technology Enhances Resilience
to Climate Change and Prevents Land Degradation:
Perceptions of Adopters in Western Kenya.” Farming
System 1 (2): 100020. https://doi.org/10.1016/j.farsys.2023.
100020 .
Neupane, D., P. Adhikari, D. Bhattarai, B. Rana, Z. Ahmed,
U. Sharma, and D. Adhikari. 2022. “Does Climate Change
Aect the Yield of the Top Three Cereals and Food
Security in the World?” Earth 3 (1): 45–71. https://doi.org/
10.3390/earth3010004 .
Nguyen, M., O. Baez-Villanueva, D. Bui, P. Nguyen, and L. Ribbe.
2020. “Harmonization of Landsat and Sentinel 2 for Crop
Monitoring in Drought Prone Areas: Case Studies of Ninh
Thuan (Vietnam) and Bekaa (Lebanon).” Remote Sensing
12 (2): 281. https://doi.org/10.3390/rs12020281 .
Ochieng, J., L. Kirimi, and M. Mathenge. 2016. “Eects of
Climate Variability and Change on Agricultural Production:
The Case of Small Scale Farmers in Kenya.” NJAS:
Wageningen Journal of Life Sciences 77 (1): 71–78. https://
doi.org/10.1016/j.njas.2016.03.005 .
Ojo, O. I., C. Temenu, and M. F. Ilunga. 2019. “Analysis of Length
of Growing Season in Some Agro-Ecological Zones Using
Non-Parametric Trend Test.” Notulae Scientia Biologicae
11 (2): 241–248. https://doi.org/10.15835/nsb11210314 .
Ondiek, R. A., M. Saber, and M. Abdel-Fattah. 2024. “Spatial–
Temporal Analysis of Impacts of Climate Variability on Maize
Yield in Kenya.” Agriculture 14 (8): 1415. https://doi.org/10.
3390/agriculture14081415 .
Otte, I., F. Detsch, E. Mwangomo, A. Hemp, T. Appelhans, and
T. Nauss. 2017. “Multidecadal Trends and Interannual
Variability of Rainfall as Observed from Five Lowland
Stations at Mt. Kilimanjaro, Tanzania.” Journal of
Hydrometeorology 18 (2): 349–361. https://doi.org/10.1175/
JHM-D-16-0062.1 .
Palmer, P. I., C. M. Wainwright, B. Dong, R. I. Maidment,
K. G. Wheeler, N. Gedney, J. E. Hickman, et al. 2023.
“Drivers and Impacts of Eastern African Rainfall Variability.”
Nature Reviews Earth and Environment 4 (4): 254–270.
https://doi.org/10.1038/s43017-023-00397-x .
Poortinga, A., K. Tenneson, A. Shapiro, Q. Nquyen, K. San Aung,
F. Chishtie, and D. Saah. 2019. “Mapping Plantations in
Myanmar by Fusing Landsat-8, Sentinel-2 and Sentinel-1
Data Along with Systematic Error Quantication.” Remote
Sensing 11 (7): 831. https://doi.org/10.3390/rs11070831 .
Qader, S. H., J. Dash, V. A. Alegana, N. R. Khwarahm, A. J. Tatem,
and P. M. Atkinson. 2021. “The Role of Earth Observation in
Achieving Sustainable Agricultural Production in Arid and
Semi-Arid Regions of the World.” Remote Sensing 13 (17):
3382. https://doi.org/10.3390/rs13173382 .
Qiu, S., Z. Zhu, and B. He. 2019. “Fmask 4.0: Improved Cloud and
Cloud Shadow Detection in Landsats 4–8 and Sentinel-2
Imagery.” Remote Sensing of Environment 231:111205.
https://doi.org/10.1016/j.rse.2019.05.024 .
Richard, K., E. M. Abdel-Rahman, S. Subramanian, J. O. Nyasani,
M. Thiel, H. Jozani, C. Borgemeister, and T. Landmann. 2017.
“Maize Cropping Systems Mapping Using RapidEye
Observations in Agro-Ecological Landscapes in Kenya.”
Sensors (Basel, Switzerland) 17 (11): 2537. https://doi.org/10.
3390/s17112537 .
Rosenzweig, C., A. Iglesius, X. B. Yang, P. Epstein, and E. Chivian.
2001. “Climate Change and Extreme Weather
Events-Implications for Food Production, Plant Diseases,
and Pests.” https://digitalcommons.unl.edu/nasapub/24/ .
Roy, D. P., J. Li, H. K. Zhang, L. Yan, H. Huang, and Z. Li. 2017.
“Examination of Sentinel-2A Multi-Spectral Instrument (MSI)
Reectance Anisotropy and the Suitability of a General
Method to Normalize MSI Reectance to Nadir BRDF
Adjusted Reectance.” Remote Sensing of Environment
199:25–38. https://doi.org/10.1016/j.rse.2017.06.019 .
Roy, D. P., H. K. Zhang, J. Ju, J. L. Gomez-Dans, P. E. Lewis,
C. B. Schaaf, Q. Sun, J. Li, H. Huang, and V. Kovalskyy. 2016.
“A General Method to Normalize Landsat Reectance Data
to Nadir BRDF Adjusted Reectance.” Remote Sensing of
Environment 176:255–271. https://doi.org/10.1016/j.rse.
2016.01.023 .
Salvo, M. D., D. Begalli, and G. Signorello. 2013. “Measuring the
Eect of Climate Change on Agriculture: A Literature Review
of Analytical Models.” Journal of Development and
Agricultural Economics 5 (12): 499–509. https://www.
researchgate.net/prole/diego-begalli/publication/
258437340_measuring_the_eect_of_climate_change_on_
agriculture_a_literature_review_of_analytical_models/
links/00b7d52838e06a79b7000000/measuring-the-eect-of
-climate-change-on-agriculture-a-literature-review-of-analy
tical-models.pdf .
Sánchez-Bermúdez, M., D. Pozo, C. Juan, and M. Pernas. 2022.
“Eects of Combined Abiotic Stresses Related to Climate
Change on Root Growth in Crops.” Frontiers in Plant
Science 13:918537. https://doi.org/10.3389/fpls.2022.
918537 .
Schlenker, W., and D. B. Lobell. 2010. “Robust Negative Impacts
of Climate Change on African Agriculture.” Environmental
Research Letters 5 (1): 14010. https://doi.org/10.1088/1748-
9326/5/1/014010 .
18 A. LIEPA ET AL.
Shahzad, A., S. Ullah, A. A. Dar, M. F. Sardar, T. Mehmood,
M. A. Tufail, A. Shakoor, and M. Haris. 2021. “Nexus on
Climate Change: Agriculture and Possible Solution to Cope
Future Climate Change Stresses.” Environmental Science and
Pollution Research International 28 (12): 14211–14232.
https://doi.org/10.1007/s11356-021-12649-8 .
Sharma, R. K., S. Kumar, K. Vatta, R. Bheemanahalli, J. Dhillon,
and K. N. Reddy. 2022. “Impact of Recent Climate Change on
Corn, Rice, and Wheat in Southeastern USA.” Scientic
Reports 12 (1): 16928. https://doi.org/10.1038/s41598-022-
21454-3 .
Sisay, B., J. Simiyu, E. Mendesil, P. Likhayo, G. Ayalew,
S. Mohamed, S. Subramanian, and T. Tefera. 2019. “Fall
Armyworm, Spodoptera Frugiperda Infestations in East
Africa: Assessment of Damage and Parasitism.” Insects
10 (7): 195. https://doi.org/10.3390/insects10070195 .
Standfuß, I., C. Geiß, R. Nathan, S. Rotics, M. Scacco, G. Kerr,
H. Taubenböck, N. Pettorelli, and G. Buchanan. 2022. “Time
Series Enable the Characterization of Small-Scale Vegetation
Dynamics That Inuence Fine-Scale Animal Behavior – An
Example from White Storks’ Foraging Behavior.” Remote
Sensing in Ecology and Conservation 8 (3): 391–408. https://
doi.org/10.1002/rse2.251 .
Stean-Dewenter, I., B. Kerr, and P. Rachel. 2024. “Insect
Diversity for Agroecosystem Resilience in a Changing
Climate.” One Earth 7 (4): 541–544. https://doi.org/10.1016/
j.oneear.2024.03.003 .
Tambo, J. A., R. K. Day, J. Lamontagne-Godwin, S. Silvestri,
P. K. Beseh, B. Oppong-Mensah, N. A. Phiri, and
M. Matimelo. 2020. “Tackling Fall Armyworm (Spodoptera
Frugiperda) Outbreak in Africa: An Analysis of farmers’
Control Actions.” International Journal of Pest Management
66 (4): 298–310. https://doi.org/10.1080/09670874.2019.
1646942 .
Taubenböck, H., J. Mast, C. Geiß, M. Wurm, T. Esch, and
K. C. Seto. 2024. “Global Dierences in Urbanization
Dynamics from 1985 to 2015 and Outlook Considering
IPCC Climate Scenarios.” Cities 151:105117. https://doi.org/
10.1016/j.cities.2024.105117 .
Torres, M., R. Howitt, and L. Rodrigues. 2019. “Analyzing Rainfall
Eects on Agricultural Income: Why Timing Matters.”
EconomiA 20 (1): 1–14. https://doi.org/10.1016/j.econ.2019.
03.006 .
Uhe, P., S. Philip, S. Kew, K. Shah, J. Kimutai, E. Mwangi, G. J. van
Oldenborgh, et al. 2018. “Attributing Drivers of the 2016
Kenyan Drought.” International Journal of Climatology
38 (S1). https://doi.org/10.1002/joc.5389 .
UN. 2018. “World Urbanization Prospects: The 2018 Revision,
Online Edition.” United Nations Department of Economic and
Social Aairs, Population Division. Accessed June, 2024.
https://population.un.org/wup/Publications/Files/
WUP2018-Report.pdf .
Vrieling, A., J. D. Leeuw, and M. Said. 2013. “Length of Growing
Period Over Africa: Variability and Trends from 30 Years of
NDVI Time Series.” Remote Sensing 5 (2): 982–1000. https://
doi.org/10.3390/rs5020982 .
Wainwright, C. M., D. L. Finney, M. Kilavi, E. Black, and
J. H. Marsham. 2021. “Extreme Rainfall in East Africa,
October 2019–January 2020 and Context Under Future
Climate Change.” Weather 76 (1): 26–31. https://doi.org/10.
1002/wea.3824 .
Wang, L., W. Zhuo, Z. Pei, X. Tong, W. Han, and S. Fang. 2021.
“Using Long-Term Earth Observation Data to Reveal the
Factors Contributing to the Early 2020 Desert Locust
Upsurge and the Resulting Vegetation Loss.” Remote
Sensing 13 (4): 680. https://doi.org/10.3390/rs13040680 .
Welch, B. L. 1947. “The Generalization of `student’s’ Problem
When Several Dierent Population Variances are Involved.”
Biometrika 34 (1/2): 28. https://doi.org/10.2307/2332510 .
White, M. A., K. M. de Beurs, K. A. M. E. L. Didan, D. W. Inouye,
A. D. Richardson, O. P. Jensen, J. O’KEEFE, et al. 2009.
“Intercomparison, Interpretation, and Assessment of Spring
Phenology in North America Estimated from Remote
Sensing for 1982–2006.” Global Change Biology 15 (10):
2335–2359. https://doi.org/10.1111/j.1365-2486.2009.
01910.x .
White, M. A., P. E. Thornton, and S. W. Running. 1997.
“A Continental Phenology Model for Monitoring
Vegetation Responses to Interannual Climatic Variability.”
Global Biogeochemical Cycles 11 (2): 217–234. https://doi.
org/10.1029/97GB00330 .
The World Bank. 2020. “Climate Risk Prole: Kenya. The World
Bank.” Accessed October 30, 2023. https://climateknowl
edgeportal.worldbank.org/sites/default/les/2021-05/
15724-WB_Kenya%20Country%20Prole-WEB.pdf .
The World Bank. 2022. “Agriculture, Forestry, and Fishing, Value
Added (% of GDP) - Kenya.” Accessed October 27, 2023.
https://data.worldbank.org/indicator/NV.AGR.TOTL.ZS?loca
tions=KE .
World Meteorological Organization. 1992. International
Meteorological Vocabulary. Vocabulaire météorologique
International. 2. ed. Genève: WMO.
Worrall, G., J. Judge, K. Boote, and A. Rangarajan. 2023. “In‐
Season Crop Phenology Using Remote Sensing and Model‐
Guided Machine Learning.” Agronomy Journal 115 (3):
1214–1236. https://doi.org/10.1002/agj2.21230 .
Yin, F., P. E. Lewis, and J. L. Gómez-Dans. 2022. “Bayesian
Atmospheric Correction Over Land: Sentinel-2/MSI and
Landsat 8/OLI.” Geoscientic Model Development 15 (21):
7933–7976. https://doi.org/10.5194/gmd-15-7933-2022 .
Zhu, Z., S. Wang, and C. E. Woodcock. 2015. “Improvement and
Expansion of the Fmask Algorithm: Cloud, Cloud Shadow,
and Snow Detection for Landsats 4–7, 8, and Sentinel 2
Images.” Remote Sensing of Environment 159:269–277.
https://doi.org/10.1016/j.rse.2014.12.014 .
Ziska, L. H., and L. L. McConnell. 2016. “Climate Change, Carbon
Dioxide, and Pest Biology: Monitor, Mitigate, Manage.”
Journal of Agricultural and Food Chemistry 64 (1): 6–12.
https://doi.org/10.1021/jf506101h.
GISCIENCE & REMOTE SENSING 19
