Field-Level Irrigation Monitoring with Integrated Use of Optical and Radar Time Series in Temperate Regions

Kód výsledku v IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F86652079%3A_____%2F22%3A00569982" target="_blank" >RIV/86652079:_____/22:00569982 - isvavai.cz</a>

Výsledek na webu

<a href="https://www.igarss2022.org/" target="_blank" >https://www.igarss2022.org/</a>

DOI - Digital Object Identifier

—

Jazyk výsledku

angličtina

Název v původním jazyce

Field-Level Irrigation Monitoring with Integrated Use of Optical and Radar Time Series in Temperate Regions

Popis výsledku v původním jazyce

Accurate and spatially explicit information on irrigation is essential for sustainable water resource management, crop condition monitoring, and modelling. Although some advances have been made for irrigation mapping using remotely sensed data, most studies are conducted in semi-arid areas, while field-level irrigation mapping remains challenging for temperate regions. To assess the applicability of different time series for irrigation mapping, we used optical and Sentinel-1 data over northern Germany. Landsat and Sentinel 2 based vegetation indices and Tasselled Cap components were aggregated over the growing season and specific key phenological stages. Irrigated areas were then classified using random forest (RF) and gradient boosting (XGboost) based classifiers. In general, XGboost outperformed RF for irrigation mapping with both composites from specific growth stages and growing seasons. Overall accuracy reached satisfactory levels (80%). The synergistic use of optical and radar data enhanced the classification accuracy.

Název v anglickém jazyce

Field-Level Irrigation Monitoring with Integrated Use of Optical and Radar Time Series in Temperate Regions

Popis výsledku anglicky

Accurate and spatially explicit information on irrigation is essential for sustainable water resource management, crop condition monitoring, and modelling. Although some advances have been made for irrigation mapping using remotely sensed data, most studies are conducted in semi-arid areas, while field-level irrigation mapping remains challenging for temperate regions. To assess the applicability of different time series for irrigation mapping, we used optical and Sentinel-1 data over northern Germany. Landsat and Sentinel 2 based vegetation indices and Tasselled Cap components were aggregated over the growing season and specific key phenological stages. Irrigated areas were then classified using random forest (RF) and gradient boosting (XGboost) based classifiers. In general, XGboost outperformed RF for irrigation mapping with both composites from specific growth stages and growing seasons. Overall accuracy reached satisfactory levels (80%). The synergistic use of optical and radar data enhanced the classification accuracy.

Druh

O - Ostatní výsledky

CEP obor

—

OECD FORD obor

20705 - Remote sensing

Projekt

—

Návaznosti

I - Institucionalni podpora na dlouhodoby koncepcni rozvoj vyzkumne organizace

Rok uplatnění

2022

Kód důvěrnosti údajů

S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů