Agro-meteorological risks to maize production in Tanzania: Sensitivity of an adapted Water Requirements Satisfaction Index (WRSI) model to rainfall

Elena Tarnavsky*, Erik Chavez, Hendrik Boogaard

*Corresponding author for this work

Research output: Contribution to journalArticleAcademicpeer-review

5 Citations (Scopus)

Abstract

The Water Requirements Satisfaction Index (WRSI) – a simplified crop water stress model – is widely used in drought and famine early warning systems, as well as in agro-meteorological risk management instruments such as crop insurance. We developed an adapted WRSI model, as introduced here, to characterise the impact of using different rainfall input datasets, ARC2, CHIRPS, and TAMSAT, on key WRSI model parameters and outputs. Results from our analyses indicate that CHIRPS best captures seasonal rainfall characteristics such as season onset and duration, which are critical for the WRSI model. Additionally, we consider planting scenarios for short-, medium-, and long-growing cycle maize and compare simulated WRSI and model outputs against reported yield at the national level for maize-growing areas in Tanzania. We find that over half of the variability in yield is explained by water stress when the CHIRPS dataset is used in the WRSI model (R2 = 0.52–0.61 for maize varieties of 120–160 days growing length). Overall, CHIRPS and TAMSAT show highest skill (R2 = 0.46–0.55 and 0.44–0.58, respectively) in capturing country-level crop yield losses related to seasonal soil moisture deficit, which is critical for drought early warning and agro-meteorological risk applications.

Original languageEnglish
Pages (from-to)77-87
JournalInternational Journal of applied Earth Observation and Geoinformation
Volume73
DOIs
Publication statusPublished - 1 Dec 2018

Keywords

  • Maize
  • Rainfall
  • Remote sensing
  • Tanzania
  • WRSI

Fingerprint Dive into the research topics of 'Agro-meteorological risks to maize production in Tanzania: Sensitivity of an adapted Water Requirements Satisfaction Index (WRSI) model to rainfall'. Together they form a unique fingerprint.

Cite this