2393-Article Text-4697-1-10-20221117
2393-Article Text-4697-1-10-20221117
Abstract
Drought is a climate change phenomenon that is difficult to avoid, so disaster mitigation planning is needed
to minimize the impact of damage. Drought potential mapping can take advantage of remote sensing data and
analysis of spatial data using a Geographic Information System (GIS). Image extraction can produce Land
Surface Temperature (LST) data, vegetation index obtained from the Normalized Difference Vegetation Index
(NDVI) transformation, land use obtained from Object-Based Image Analysis (OBIA), and wetness index from
the Normalized Difference Water Index (NDWI). This study integrates data between image extraction results
and regional conditions such as rainfall, geological, soil types, and hydrogeology. Klaten Regency has the
potential for very high-class drought covering an area of 101.53 ha. In Bayat district, the results of the
identification of potential drought indicate very high levels of drought.
1. Introduction
Indonesia is located on the equator, so there is a rainy detrimental natural disaster. Therefore, drought has the
and dry season throughout the year. In the dry season, most damaging impact on various sectors compared to
drought often occurs in several areas due to the other natural disasters because it causes water scarcity,
duration of the dry season being too long. Drought is a agricultural drought, and famine (Smith and Katz,
natural disaster that is difficult to predict but can be 2013). Each region has different characteristics so it is
detrimental to human life. Indonesia often experiences difficult to determine the approach to identifying
meteorological drought in several areas. accurate drought characteristics (Hao et al., 2017).
Meteorological drought variables are based on rainfall Difficulties in identifying droughts encourage
and temperature whose application can be determined researchers to develop related indicators of drought
on a local or regional scale depending on the including the applications used, regional conditions,
availability of data and the spatial distribution of the and data availability in the area. In general, researchers
earth station network (Rhee et al., 2010). The focus their research on drought in certain geographic
availability of data makes it difficult to predict drought areas that have the potential for drought (Zhang et al.,
in large areas. Drought is caused by an uneven 2017a). Drought mapping has temporal and spatial
distribution of rainfall over a long period in an area complexity, making it difficult to accurately determine
(Lei and Duan, 2011 and Jamil et al., 2013). Drought and identify the start and end of the drought and
can also be interpreted as a lack of water supply drought duration (Wu et al., 2013). The potential for
compared to the need for water for human purposes. drought requires accurate spatial data in describing
Drought affects various sectors in society such as information about an area so that it can plan for
agriculture, ecosystem services, human health, handling drought disasters if at any time the disaster
recreation, and water resources and is the most occurs.
Geologi
Landsat 8 Sentinel-2 DEMNAS Hidrogeology Rainfall Field Survey
Soil Types
Preprocessing Preprocessing Preprocessing Metode
Poligon
Thiessen
LST NDWI NDVI OBIA Slope Map Classification
Map of Map of
Land Use Rainfall
Scoring
Validation
Drought
Map
Figure 2 shows the flow of research carried out to The field survey is used to validate the results of
identify drought in Klaten Regency. Meanwhile, the drought identification and land use maps. Research
research data can be seen in Table 1 and Figure 3. data using Sentinel 2 L1C recording on September
Landsat 8 data were processed to obtain LST, NDWI, 26, 2019 in the dry season. Data collection in the dry
and NDVI values. Then, Citra Sentinel 2 produces a season aims to determine the effect of the vegetation
land use map from the results of OBIA processing. index on the potential for drought. The dry season
Slope Map obtained from DEMNAS. Regional causes plants to become dry and rice fields become
condition data such as geological, hydrogeological, unproductive. Vegetation growth is influenced by
soil type, and map of rainfall data are classified for monsoon rainfall so that the health status of
drought analysis. All data were scored to identify the vegetation is closely related to rainfall (Ramadan et
drought disaster in Klaten Regency. al., 2021).
2.1 Object-Based Image Analysis (OBIA) segmentation, the process divides the entire image area
The OBIA algorithm extracts satellite image data into into several non-overlapping polygonal object shapes
an object based on its spectral, spatial, and texture according to homogeneity or heterogeneity criteria,
characteristics (Hay and Castilla, 2008). OBIA and each object has the same internal feature shape
produces satellite image classification by grouping including spectrum, spatial, texture, shape, etc.
homogeneous pixels based on the identification of the (Dilpreet and Yadwinder, 2014).
similarity of spectral values in the formation of an
object delineation by utilizing elements such as 2.2 Normalized Difference Vegetation Index (NDVI)
spectral, texture, and spatial. OBIA has advantages Multispectral satellite imagery data is extracted to
compared to traditional pixel-based classification obtain the vegetation density of an area easily and
methods (eg maximum likelihood classifier) because quickly (Pirotti et al., 2014). Vegetation density is
this technique performs grouping of the same extracted from satellite imagery based on estimates of
neighboring pixels into a delineation object using photosynthetic activity in a vegetated area by
specified parameters (Hulet et al., 2014). OBIA also combining the Red(R) and Near-Infrared (NIR) bands
identifies objects based on the pattern recognition which can show the presence of chlorophyll (Spadoni
process (Lei and Duan, 2011). OBIA utilizes a bottom- et al., 2020). The two bands are used as vegetation
up approach by first detecting the spectral values for index parameters because the results of the band size
each pixel that are homogeneous in the formation of are influenced by chlorophyll absorption, are sensitive
small objects, then combining them into large objects to vegetation biomass, and can distinguish between
according to their actual shape. The OBIA algorithm vegetated land, open land, and water. The results of
extracts satellite image data based on image NDVI processing get a comparison of the ratio values.
Table 3 and Figure 5 show the results of NDVI center has a low vegetation density due to a lot of
processing in the Klaten Regency. These results build-up land in the area. Low vegetation density is
indicate that the vegetation density in the study area also found in the southern area because it is located
is mostly in the medium to high vegetation density at the foot of the low hills, namely the Jiwo Hills. The
category. The NDVI results can show that most of the plain is composed of Quaternary deposits. The rock
Klaten Regency area is an agricultural land area that unit consists of alternating sandstone and siltstone
is by the land use classification. The category of low and limestone lenses. This causes the area to have
vegetation density is found in the city center. The city low-density vegetation found in Bayat district.
Therefore, the slope data are classified in the range of The scores for each parameter were summed so that
0-8%, 5-15%, 15-25%, 25-45%, and >45%. Klaten the scores ranged from 15 to 38. The scores were
Regency has a diversity of geological structures that divided into 5 classes with the same distribution.
can provide different potential impacts on each type Figure 8 shows the identification of drought in Klaten
of drought. Geological structures are classified based Regency. The majority of the Klaten Regency area
on their types that can affect the potential for drought does not have the potential for drought, but there are
in Klaten Regency. Geological structure affects some areas that have been identified as drought.
surface water flow. Soil type greatly affects drought Based on the results of the processing, several sub-
because it affects the ability of the soil to hold water districts have the potential for drought including the
below the surface. The ability of the soil is districts of Kemalang, Manisrenggo, Jatinom,
represented by the condition of soil moisture in Prambanan, Trucuk, Bayat, Pedan, and Karangdowo.
storing water so that it can indicate the level of The results of the classification show that Bayat
drought based on the availability of groundwater. district has very high, high, and moderate drought
Klaten Regency consists of soil types in the form of classes, indicating that the area has a high potential
gray regosol, gray brown regosol, regosol dark grey, for drought in the dry season. The Bayat district area
grumusol dark grey, and litosol which are classified has the potential to experience drought due to the
based on the ability of the soil to hold water below type of soil in the form of littoral soil. This type of
the surface. All parameters are done by adding up the soil is a type of soil that is difficult for plants to grow
scores for each predetermined class. The scoring because of its low nutrient content. Hydrogeological
values for all parameters are added up to obtain a data shows Bayat district is an area where
range of values between 15 to 38. The results of the groundwater is scarce. This causes Bayat district to
range of potential drought values are classified based have a high potential for drought. Bayat district has
on the level of drought which can be seen in Table 6 an annual rainfall volume with a relatively moderate
by dividing into 5 classes with the same range value category of 2,073 mm/year obtained from Bawak
for each class. station.