RUSSIAN JOURNAL OF EARTH SCIENCES, VOL. 19, ES2003, doi:10.2205/2018ES000647, 2019
Le Hung Trinh1, Danh Tuyen Vu2
1Le Quy Don Technical University, Hanoi, Vietnam
2Hanoi University of Natural Resources and Environment, Hanoi, Vietnam
Drought is a natural phenomenon, which occurs in most regions in the world, caused immense damage in agricultural production and seriously affected on the environment. Droughts often occur on a large scale, so traditional observation techniques for drought monitoring are difficult, time consuming and costly. Remote sensing data provides information on the Earth's surface at different spectrum channels and and wide coverage has been used effectively in monitoring and severe drought. This paper presents the results of the study on drought assessment in Bac Binh district, Binh Thuan province using LANDSAT multispectral imagery based on normalized difference drought index (NDDI). The results obtained in the study can be used to create drought level map and provides information to help managers set measures to minimize damage caused by drought.
Drought is one of the most costly natural disasters in Vietnam. In recent years, drought tends to happen both in rainy season and dry season due to the impact of climate change [Ngo et al., 2017; Pham and Nguyen, 2012; Trinh, 2014]. Droughts are particularly severe in the Central and Central Highlands of Vietnam that causes great damage in the socio-economic and environment sectors. According to the Ministry of Agriculture and Rural Development Vietnam, more than 240,000 hectares of rice, 18,000 hectares of crops, 55,000 hectares of fruit trees and more than 100,000 hectares of industrial crops were damaged by water shortages only in drought years 2015–2016 [Ministry..., 2008].
The traditional drought monitoring methods used in Vietnam are based on data on precipitation, temperature, and soil moisture collected from the meteorological stations. Ground-based observations reflect only drought situation of local area around the station and in fact cannot establish the number of meteorological stations with expected density due to the high cost. Therefore, such data-sets fail to monitor drought conditions accurately and in a timely fashion. Remote sensing technology with advantages such as wide area coverage and short revisit interval has been used effectively in the study of drought monitoring.
Optical remote sensing data such as MODIS, LANDSAT and SPOT have been used for the large-scale drought monitoring. During the past few years, many studies in Vietnam have demonstrated the effectiveness of remote sensing techniques for drought monitoring and assessment of soil/vegetation moisture [Ngo et al., 2017; Nguyen et al., 2016; Tran, 2007; Trinh, 2014].
More than 100 drought indices have so far been proposed, some of which are operationally used to characterize drought using gridded maps at regional and national levels. These indices correspond to different types of drought, including meteorological, agricultural, and hydrological drought [Gulacsi and Kovacs, 2015; Zagar et al., 2011]. Some of the indices derived from satellite data and used for drought monitoring include vegetation condition index – VCI [Kogan, 1990], crop water stress index – CWSI [Idso et al., 1981], normalized difference infrared index – NDII [Hardisky et al., 1983], vegetation health index – VHI [Kogan, 1995], normalized difference water index – NDWI [Gao, 1996], global vegetation moisture index – GVMI [Ceccato et al., 2002], soil water index – SWI [Wagner, 2003], remote sensing drought risk index – RDRI [Liu et al., 2008]. Some drought indices are based on the relationship between land surface temperature and land cover, such as the temperature condition index – TCI [Kogan, 1995], vegetation temperature condition index – VTCI [Wang et al., 2001], temperature vegetation dryness index – TVDI [Lambin and Ehrlich, 1996; Sandholt et al., 2002].
Gu et al. [2007] proposed the normalized difference drought index (NDDI), which combines both vegetation greenness (NDVI) [Rouse et al., 1974] and wetness conditions (NDWI) [Gao, 1996]. This drought index can be suitable for long-term drought monitoring, particularly for agricultural drought [Hazaymeh and Hassan, 2016].
This paper focuses on the application of normalized difference drought index to assess and monitor drought in Bac Binh district, Binh Thuan province (South Central Coast of Vietnam) in period 1988–2017.
|
| Figure 1 |
Bac Binh is a rural district of Binh Thuan province in the South Central Coast region of Vietnam (Figure 1). The area is bounded between latitude 10° $58'27"$ N to 11° $31'38"$ N and longitude 108° $06'30"$ E to 108° $37'34"$ E. The district covers an area of 1868.82 km$^2$ and had a population over of 120,000 people (http://bacbinh.binhthuan.gov.vn).
Bac Binh has three geographical terrains: plain, semi-mountain and mountain. It located in tropical monsoon region and is characterized by semi-arid climate in the south central region. The climate is divided into two seasons: the rainy season from May until October and the dry season from November to April. According the monitoring results, the average annual rainfall in Bac Binh district is very low and the average temperature is high. This is one of the most regions severely affected by drought in Southeast region of Vietnam. The major form of desertification in Bac Binh district is desertification of sand dunes. The total red sandy area is 32,600 ha (17.5% total area of district).
 |
| Figure 2 |
 |
| Table 1 |
In this study, nine multispectral cloud – free LANDSAT 5 TM, LANDSAT 7 ETM$+$ and 8 OLI-TIRS images (path 124, row 52) with a spatial resolution of $30\times30$ meters were acquired from 1988, 1993, 1996, 1998, 2002, 2004, 2009, 2011 and 2017 (Figure 2 and Table 1). The LANDSAT data was the standard terrain correction products (L1T), downloaded from United States Geological Survey (USGS – http://glovis.usgs.gov) website. These satellite images were taken during the peak time of the dry season in South Central Coast of Vietnam (January to March).
Image processing started with radiometric and geometric correction. In first step, the brightness of the pixel value (digital number) is converted into the spectral radiance value (Wm$^{-2}$ $\mu$m$^{-1}$). Then, the spectral radiance value is used to calculate the top of atmosphere value – TOA. The surface reflectance value can be calculated using "dark object subtraction" (DOS) atmospheric correction method [Chavez, 1988, 1996]. This method estimates the atmospheric contributions to a surface spectrum by measuring homogeneous surfaces over a range of illumination conditions.
The reflectance values of green, red, near infrared (NIR) and short wave infrared (SWIR) bands were used to calculate normalized difference vegetation index (NDVI) [Rouse et al., 1974] and normalized difference water index [McFeeters, 1996] following the equations:
\begin{eqnarray*} \mathrm{NDVI} = \frac{\rho_{\mathrm{NIR}} - \rho_{\mathrm{RED}}} {\rho_{\mathrm{NIR}} + \rho_{\mathrm{RED}}} \end{eqnarray*} \begin{eqnarray*} \mathrm{NDWI} = \frac{\rho_{\mathrm{GREEN}} - \rho_{\mathrm{SWIR}}} {\rho_{\mathrm{GREEN}} + \rho_{\mathrm{SWIR}}} \end{eqnarray*}where $\rho_{\mathrm{GREEN}}$, $\rho_{\mathrm{RED}}$, $\rho_{\mathrm{NIR}}$, $\rho_{\mathrm{SWIR}}$ are reflectance values of green, red, near infrared (NIR) and short wave infrared (SWIR) bands of LANDSAT multispectral image.
Finally, the NDDI index combines information from both the NDVI and NDWI indices following the equations [Gu et al., 2007]:
\begin{equation} \tag*{(1)} \mathrm{NDDI} = \frac{\mathrm{NDVI} - \mathrm{NDWI}} {\mathrm{NDVI} + \mathrm{NDWI}} \end{equation} |
| Table 2 |
According to Drought Categories [Gu et al., 2007], NDDI index was classified into 3 classes: "non-drought" (NDDI value is smaller than 0.1); "drought-moderate" (NDDI value range from 0.1 to 0.3) and "drought-severe" (NDDI value larger than 0.3) (Table 2).
In this study, image processing is done by using ERDAS Imagine 2014 program, and drought maps were created using ArcGIS 10 program.
 |
| Figure 3 |
 |
| Figure 4 |
 |
| Table 3 |
The LANDSAT multi-temporal data after pre-processing was cut along the boundary of study area. The reflectance values for red and NIR bands were used to calculate normalized difference vegetation index and the reflectance values of green and SWIR bands were used to calculate normalized difference water index. Subsequently, the NDVI and NDWI indices are used to calculate the normalized difference drought index following formula (1). The result of NDDI index in Bac Binh district retrieval from LANDSAT satellite images in period 1988–2017 is present in Figure 3. The dark pixels of NDDI index have low drought level, and the light pixels have high drought level. The inter-annual variation in the percentage of drought impacted areas in the Bac Binh district from 1988–2017 due to drought categories [Gu et al., 2007] are shown in Figure 4 and Table 3.
Thus, it can be seen that relative large drought areas recorded in the middle of the 1993–2002 period, in which the area affected by drought severe was highest in 1996 (13.02% of total area) and 1998 (18.78% of total area) (Table 3). This is also the time when the south central coast region of Vietnam is severely affected by the El Nino phenomenon (El Nino years in period 1993–1995 and 1997–1998).
Analysis of the obtained results showed that, the regions with "non-drought" level are mainly concentrated in forested areas, where distributed in the north and a part in the south of the study area (vegetation cover is shown by red color on the Figure 2). The areas with "non-drought" level strongly decreased in period 1988–1998, from 80.17% of total area to 49.50% of total area due to the decline of forest area, especially in the northwest region of Bac Binh district. These areas are likely to increase in period 1998–2009, from 49.50% to 70.04% of total area of district and then continue to decrease in 2011 (65.41% of total area) and 2017 years (58.67% of total area).
The results obtained in this study also show the important role of vegetation cover in reducing drought risk. In planted forest areas (Phan Hiep, Hoa Thang, Hong Thai regions) in east of the district, drought level is at "severe" in period 1988–1996 to "moderate" and "non-drought" in 2011 and 2017 years (Figure 4).
 |
| Figure 5 |
 |
| Table 4 |
Measurement data of soil moisture and land surface temperature at 10 sample points (Figure 5) at 23 February 2017 were used in the comparison with NDDI values retrieval from LANDSAT 8 image, which acquired from same day. This in situ data was observed in the framework of the ministry-level project (Ministry of Natural Resources and Environment (Vietnam), No. 2015.08.10). Comparison between NDDI values and measurement data of soil moisture and land surface temperature is presented in Table 4. It can be seen, 8 out of 10 measurement points are located at "drought-severe" level and 2 measurement points are located at "drought- moderate" level. The soil moisture at the all 10 sampling points is less than 10%, and land surface temperature is high than 305 (K). The soil moisture at sampling points at "drought-moderate" level is highest, and the land surface temperature at these points is lowest in comparison with the measurement points with drought level at "severe".
We also compared the results obtained in the study with the meteorology index (MI) values which calculated from rainfall data at 13 meteorological stations in the period 1960–2010 Pham and Nguyen, 2012]. These meteorological stations are located in Bac Binh district and some neighboring districts in Binh Thuan province. The results showed that the northern part of study area has drought level "non-drought" and "less drought" (the MI values higher than 1.2). Meanwhile, the central and southern part of Bac Binh district suffered severe drought (the MI values less than 0.4). This result is also consistent with the distribution map of drought in the Bac Binh district (Figure 4).
In this study, we used Landsat time series data in period 1988–2017 to assess drought conditions in Bac Binh district, Binh Thuan province (South Central Coast of Vietnam). The results obtained show that NDDI index can be used effectively for drought monitoring due to simplicity in calculation.
The results show that relative large drought areas recorded in the middle of the 1993–2002 period, in which the area affected by drought severe was highest in 1996 and 1998 (En Nino years). The drought-prone areas are mainly distributed in agricultural and sandy lands, which is concentrated in the center and southern part of the study area. In spite of the large inter-annual variability, the long-term trends in the drought impacted areas in the Bac Binh district as a whole increased dramatically over the period 1988–2017, especially in the El Nino years of 1993, 1996 and 1998.
The results obtained in the study can be used to create drought distribution map and provide information to effectively respond and minimize damage caused by drought.
Ceccato, P., N. Gobron, S. Flasse, B. Pinty, S. Taratola (2002), Designing a spectral index to estimate vegetation content from remote sensing data: Part 1: Theoretical approach, Remote Sensing of Environment, 82, no. 2–3, p. 188–197, https://doi.org/10.1016/S0034-4257(02)00037-8.
Chavez, P. S. (1988), An improved dark-object subtraction technique for atmospheric scattering correction of multispectral data, Remote Sensing of Environment, 24, p. 459–479, https://doi.org/10.1016/0034-4257(88)90019-3.
Chavez, P. S. (1996), Image-based atmospheric corrections–revisited and improved, Photogrammetric Engineering and Remote Sensing, 62, no. 9, p. 1025–1036.
Gao, B. C. (1996), NDWI – A normalized difference water index for remote sensing of vegetation liquid water from space, Remote Sensing of Environment, 58, p. 257–266, https://doi.org/10.1016/S0034-4257(96)00067-3.
Gulacsi, A., F. Kovacs (2015), Drought monitoring with spectral indices calculated from MODIS satellite images in Hungary, Journal of Environmental Geography, 8, no. 3–4, p. 11–20, https://doi.org/10.1515/jengeo-2015-0008.
Gu, Y., J. F. Brown, J. P. Verdin, B. Wardlow (2007), A five year analysis of MODIS NDVI and NDWI for grassland drought assessment over the central Great Plains of the United States, Geophysiscal Research Letters, 34, no. 6, p. 1–6, https://doi.org/10.1029/2006GL029127.
Hardisky, M., V. Klemas, R. Smart (1983), The influence of soil salinity, growth form and leaf moisture on the spectral radiance of Spartina alterniflora canopies, Photogrammetric Engineering and Remote Sensing, 49, p. 77–83.
Hazaymed, K., Q. Hassan (2016), Remote sensing of agricultural drought monitoring: a state of art review, Special issue in the journal Remote Sensing: "Remote Sensing of Wildfire", 3, no. 4, p. 604–630.
Idso, S., R. Jackson, P. Pinter, R. Reginato, J. Hatfield (1981), Normalizing the stress-degree-day parameter for environmental variability, Agricultural Meteorology, 24, p. 45–55, https://doi.org/10.1016/0002-1571(81)90032-7.
Kogan, F. N. (1990), Remote sensing of weather impacts on vegetation in non-homogeneous areas, International Journal of Remote Sensing, 11, no. 8, p. 1405–1419, https://doi.org/10.1080/01431169008955102.
Kogan, F. N. (1995), Application of vegetation index and brightness temperature for drought detection, Advances of Space Research, 15, no. 11, p. 91–100, https://doi.org/10.1016/0273-1177(95)00079-T.
Lambin, T. R., D. Ehrlich (1996), The surface temperature-vegetation index space for land cover and land cover change analysis, International Journal of Remote Sensing, 17, no. 3, p. 163–187, https://doi.org/10.1080/01431169608949021.
Liu, L., D. Xiang, Z. Zhou, Dong (2008), Analyses the modifications of the drought monitoring model based on the cloud parameters method, 2008 Congress on Image and Signal Processing, p. 687–691, Sanya, Hainan, China, https://doi.org/10.1109/CISP.2008.431.
McFeeters, S. K. (1996), The use of the normalized difference water index (NDWI) in the delineation open water features, International Journal of Remote Sensing, 17, p. 1425–1432, https://doi.org/10.1080/01431169608948714.
Ministry of Agriculture and Rural Development, (2008), Survey project for desertification in central area from Khanh Hoa to Ninh Thuan provinces, 75 pp., Ministry of Agriculture and Rural Development, Vietnam.
Ngo, T. D., T. T. H. Nguyen, T. P. T. Nguyen, et al. (2017), 30 years monitoring spatial-temporal dynamics of agricultural drought in the central highlands using LANDSAT data, Geo-spatial Technologies and Earth Resources (GTER 2017), p. 181–188, Publishing House for Science and Technology, Hanoi, Vietnam.
Nguyen, T. T. H., et al. (2016), Mapping droughts over the central highland of Vietnam in El Nino years using LANDSAT imageries, Vietnam National University Journal of Earth Sciences, 26, p. 75–81.
Pham, Q. V., T. T. H. Nguyen (2012), Assessing agricultural drought for Binh Thuan province under climate change scenario, Vietnam Journal of Earth Sciences, 34, no. 4, p. 513–523.
Rouse, J. W., H. R. Haas, A. J. Schell, W. D. Deering (1974), Monitoring vegetation systems in the Great Plains with ERTS, Third ERTS Symposium, NASA SP-351, 1, p. 309–317.
Sandholt, I., K. Rasmussen, J. Anderson (2002), A simple interpretation of the surface temperature/vegetation index space for assessment of the surface moisture status, Remote Sensing of Environment, 79, p. 213–224, https://doi.org/10.1016/S0034-4257(01)00274-7.
Tran, H. (2007), MODIS data to monitor surface soil/vegetation moisture status: a study with temperature vegetation dryness index, Vietnam Journal of Remote Sensing and Geomatics, 2, p. 38–45.
Trinh, L. H. (2014), Application of LANDSAT thermal infrared data to study soil moisture using temperature vegetation dryness index, Vietnam Journal of Earth Sciences, 36, no. 3, p. 262–270.
Wagner, W., K. Scipal, C. Pathe, D. Gerten, W. Lucht, B. Rudolf (2003), Evaluation of the agreement between the first global remotely sensed soil moisture data with model and precipitation data, Journal of Geography Research, 108, no. D19, p. 4611, https://doi.org/10.1029/2003JD003663.
Wang, L., X. Li, J. Gong, C. Song (2001), Vegetation temperature condition index and its application for drought monitoring, Geoscience and Remote Sensing Symposium, IGARSS'01, p. 141–143, IEEE International Institute of Electrical and Electronics Engineers, Fort Worth, Texas, USA.
Zagar, A., R. Sadiq, B. Naser, F. I. Khan (2011), A review of drought indices, Environment Review, 19, p. 333–349, https://doi.org/10.1139/a11-013.
Received 27 September 2018; accepted 21 December 2018; published 14 February 2019.
Citation: Trinh Le Hung, Danh Tuyen Vu (2019), Application of remote sensing technique for drought assessment based on normalized difference drought index, a case study of Bac Binh district, Binh Thuan province (Vietnam), Russ. J. Earth Sci., 19, ES2003, doi:10.2205/2018ES000647.
Copyright 2019 by the Geophysical Center RAS.