Estimation of Green Water Footprint of Rice Paddies in taitung area using MODIS data

Chih Da Wu; Tzu Yu Huang; Shih Yuan Lin; Chun Te Lin

Estimation of Green Water Footprint of Rice Paddies in taitung area using MODIS data

Chih Da Wu, Tzu Yu Huang, Shih Yuan Lin, Chun Te Lin

Department of Geomatics

Research output: Contribution to conference › Paper › peer-review

Abstract

Green Water Footprint (GWF) is a recently developed indicator to identify the utilization and availability of the fresh water resource provided for agricultural products. Based on it accounting for rainwater evapotranspiration, a MODIS Global Terrestrial Evapotranspiration Data Set (MOD16) is applied to estimate GWF for its capable presenting global evapotranspiration status with high accuracy, wide coverage and long-term monitoring. Although the MOD16 product offers aforementioned advantages, current drawback is mainly on its data available time is too slow, for which announced data is about one-year late. This paper therefore aims to overcome the drawbacks and develop a regression method considering multiple variables including weather parameters and Normalized Difference Vegetation Index (NDVI) values in order to improve that the current MOD16 cannot reflect a more near real-time GWF. To demonstrate feasibility of the method we proposed, three representative agricultural areas in Taitung County mainly used for rice planting were selected as the studied sites. The analyzing data covered a prolonged period (from 2003 to 2012) and our regression model further distinguished the first and second cropping seasons in the ten years. The results of stepwise regression analysis reported that temperature and NDVI were significant variables related to MOD16. R² values of models derived from the 1^st and 2^nd cropping data were 0.65 and 0.64, respectively. The regression models were also verified by 10 years' out-of-samples, and the results indicated that overall accuracy of the prediction was above 85%. Since the modelled evapotranspiration value was reliable, it was then used to compute the rice green water footprint of the two cropping seasons. Based on the model, even without values of the latest MOD16, the GWF of rice producing over the same area in 2013 and 2014 were estimated. The potential of a near real-time estimation of GWF of rice was demonstrated.

Original language	English
Publication status	Published - 2015
Event	36th Asian Conference on Remote Sensing: Fostering Resilient Growth in Asia, ACRS 2015 - Quezon City, Metro Manila, Philippines Duration: 2015 Oct 24 → 2015 Oct 28

Other

Other	36th Asian Conference on Remote Sensing: Fostering Resilient Growth in Asia, ACRS 2015
Country/Territory	Philippines
City	Quezon City, Metro Manila
Period	15-10-24 → 15-10-28

All Science Journal Classification (ASJC) codes

Computer Networks and Communications

Cite this

@conference{425d837b7c874ba4ab39f2d8fdc22d9c,

title = "Estimation of Green Water Footprint of Rice Paddies in taitung area using MODIS data",

abstract = "Green Water Footprint (GWF) is a recently developed indicator to identify the utilization and availability of the fresh water resource provided for agricultural products. Based on it accounting for rainwater evapotranspiration, a MODIS Global Terrestrial Evapotranspiration Data Set (MOD16) is applied to estimate GWF for its capable presenting global evapotranspiration status with high accuracy, wide coverage and long-term monitoring. Although the MOD16 product offers aforementioned advantages, current drawback is mainly on its data available time is too slow, for which announced data is about one-year late. This paper therefore aims to overcome the drawbacks and develop a regression method considering multiple variables including weather parameters and Normalized Difference Vegetation Index (NDVI) values in order to improve that the current MOD16 cannot reflect a more near real-time GWF. To demonstrate feasibility of the method we proposed, three representative agricultural areas in Taitung County mainly used for rice planting were selected as the studied sites. The analyzing data covered a prolonged period (from 2003 to 2012) and our regression model further distinguished the first and second cropping seasons in the ten years. The results of stepwise regression analysis reported that temperature and NDVI were significant variables related to MOD16. R2 values of models derived from the 1st and 2nd cropping data were 0.65 and 0.64, respectively. The regression models were also verified by 10 years' out-of-samples, and the results indicated that overall accuracy of the prediction was above 85%. Since the modelled evapotranspiration value was reliable, it was then used to compute the rice green water footprint of the two cropping seasons. Based on the model, even without values of the latest MOD16, the GWF of rice producing over the same area in 2013 and 2014 were estimated. The potential of a near real-time estimation of GWF of rice was demonstrated.",

author = "Wu, {Chih Da} and Huang, {Tzu Yu} and Lin, {Shih Yuan} and Lin, {Chun Te}",

year = "2015",

language = "English",

note = "36th Asian Conference on Remote Sensing: Fostering Resilient Growth in Asia, ACRS 2015 ; Conference date: 24-10-2015 Through 28-10-2015",

}

TY - CONF

T1 - Estimation of Green Water Footprint of Rice Paddies in taitung area using MODIS data

AU - Wu, Chih Da

AU - Huang, Tzu Yu

AU - Lin, Shih Yuan

AU - Lin, Chun Te

PY - 2015

Y1 - 2015

N2 - Green Water Footprint (GWF) is a recently developed indicator to identify the utilization and availability of the fresh water resource provided for agricultural products. Based on it accounting for rainwater evapotranspiration, a MODIS Global Terrestrial Evapotranspiration Data Set (MOD16) is applied to estimate GWF for its capable presenting global evapotranspiration status with high accuracy, wide coverage and long-term monitoring. Although the MOD16 product offers aforementioned advantages, current drawback is mainly on its data available time is too slow, for which announced data is about one-year late. This paper therefore aims to overcome the drawbacks and develop a regression method considering multiple variables including weather parameters and Normalized Difference Vegetation Index (NDVI) values in order to improve that the current MOD16 cannot reflect a more near real-time GWF. To demonstrate feasibility of the method we proposed, three representative agricultural areas in Taitung County mainly used for rice planting were selected as the studied sites. The analyzing data covered a prolonged period (from 2003 to 2012) and our regression model further distinguished the first and second cropping seasons in the ten years. The results of stepwise regression analysis reported that temperature and NDVI were significant variables related to MOD16. R2 values of models derived from the 1st and 2nd cropping data were 0.65 and 0.64, respectively. The regression models were also verified by 10 years' out-of-samples, and the results indicated that overall accuracy of the prediction was above 85%. Since the modelled evapotranspiration value was reliable, it was then used to compute the rice green water footprint of the two cropping seasons. Based on the model, even without values of the latest MOD16, the GWF of rice producing over the same area in 2013 and 2014 were estimated. The potential of a near real-time estimation of GWF of rice was demonstrated.

AB - Green Water Footprint (GWF) is a recently developed indicator to identify the utilization and availability of the fresh water resource provided for agricultural products. Based on it accounting for rainwater evapotranspiration, a MODIS Global Terrestrial Evapotranspiration Data Set (MOD16) is applied to estimate GWF for its capable presenting global evapotranspiration status with high accuracy, wide coverage and long-term monitoring. Although the MOD16 product offers aforementioned advantages, current drawback is mainly on its data available time is too slow, for which announced data is about one-year late. This paper therefore aims to overcome the drawbacks and develop a regression method considering multiple variables including weather parameters and Normalized Difference Vegetation Index (NDVI) values in order to improve that the current MOD16 cannot reflect a more near real-time GWF. To demonstrate feasibility of the method we proposed, three representative agricultural areas in Taitung County mainly used for rice planting were selected as the studied sites. The analyzing data covered a prolonged period (from 2003 to 2012) and our regression model further distinguished the first and second cropping seasons in the ten years. The results of stepwise regression analysis reported that temperature and NDVI were significant variables related to MOD16. R2 values of models derived from the 1st and 2nd cropping data were 0.65 and 0.64, respectively. The regression models were also verified by 10 years' out-of-samples, and the results indicated that overall accuracy of the prediction was above 85%. Since the modelled evapotranspiration value was reliable, it was then used to compute the rice green water footprint of the two cropping seasons. Based on the model, even without values of the latest MOD16, the GWF of rice producing over the same area in 2013 and 2014 were estimated. The potential of a near real-time estimation of GWF of rice was demonstrated.

UR - http://www.scopus.com/inward/record.url?scp=84964039870&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84964039870&partnerID=8YFLogxK

M3 - Paper

AN - SCOPUS:84964039870

T2 - 36th Asian Conference on Remote Sensing: Fostering Resilient Growth in Asia, ACRS 2015

Y2 - 24 October 2015 through 28 October 2015

ER -

Estimation of Green Water Footprint of Rice Paddies in taitung area using MODIS data

Abstract

Other

All Science Journal Classification (ASJC) codes

Other files and links

Fingerprint

Cite this