Investigation of the Variability of Dry / Wet Meteorological conditions using the SPEI / SPI Index in the Selo Watershed, Period 1981-2020.

The significant impact of climate change is high variability in rainfall and an increase in extreme rainfall events. The impacts resulted in severe dry/wet conditions. Severe dry conditions can lead to crop failure due to lack of water availability, while severe wet conditions can indicate flooding. This paper aims to investigate the variability of dry/wet meteorological conditions in the Selo watershed. The investigation was carried out using the SPEI and SPI. The results show that wet/dry conditions do not have a clear pattern variability in the long term. Seasonal wet/dry conditions show a changing trend towards wetter conditions in all seasons. Spatially, the changes occur throughout the study area, with the most significant changes occurring in the upstream and downstream parts of the watershed. In the last two decades, wet conditions are more dominant to replace dry conditions that were more prominent in the previous two decades. Other analysis results conclude that there is no influence of air temperature in creating wet/dry conditions in the Selo watershed area. So that the characteristics of wet/dry conditions are very useful for considering the selection of agricultural adaptation strategies in dealing with variability and climate change in the Selo watershed area.


Introduction
Climate change significantly affects the hydrological cycle due to changes in the spatial and temporal distribution of water resources.Changes in the distribution of water resources will cause drought if there is a shortage of water and flooding if there is excess water.The variability of wet/dry conditions greatly affects agricultural productivity because lack of water can develop into a drought hazard if it lasts for a long time (several months), whereas if there is excess water in a short time, it will cause flooding.Obviously, these two phenomena will cause crop failure.Floods and droughts are definitely marked by a surplus (wet) or water deficit (dry) in an environment.Until now, various drought indices have been developed to assess wet/dry environmental conditions.Drought index models that can be used in assessing the impact of climate change on drought are drought indices developed based on the cumulative amount of rainfall and evapotranspiration because rainfall and evapotranspiration are directly affected by climate change [1,2].Operationally, drought indices use rainfall to quantify drought 1306 (2024) 012043 IOP Publishing doi:10.1088/1755-1315/1306/1/012043 2 characteristics, either singly or in combination with other meteorological parameters, depending on the model requirements and the purpose of making and or developing the drought index (meteorology, agriculture and hydrology) [3,4].
Many research results show an increase in the frequency and magnitude of extreme climate events that have occurred in the last few decades, which have an impact on the frequent occurrence of droughts and floods.It becomes very important to investigate the variability and tendency of wet/dry conditions in a watershed, given its role as a key factor in the hydrological cycle [5], [6].Therefore, the purpose of this study is to characterize wet/dry conditions in the selection of adaptation strategies for the availability of water resources due to climate variability in the Selo watershed, Tanah Datar Regency.

Methodology 2.1. Research sites
The research is planned to be carried out in the Selo watershed area, which is in the administrative area of Tanah Datar Regency, covering seven sub-districts in the district: Salimpaung, Sungayang, Sungai Tarab, Pariangan, Lima Kaum, Rambatan, Padang Ganting and Tanjung Emas sub-districts.The Selo watershed is a catchment and drain of water from the Selo river, which originates on the east-southeast slope of Mount Marapi and southwest of the slopes of Mount Sago and empties into Batang Sinamar in Tanjung Emas District.In general, Batang Selo has a dendritic flow pattern with the direction of flow to the southeast from its upstream direction, with denser tributaries in the upstream west (Figure 1).Climatologically, the amount of rainfall in the Selo watershed is 2000 mm/year.Its location in the shadow of rain causes the watershed to get relatively less rainfall when compared to the western coast of West Sumatra.SRTM imagery and Peta Rupa Bumi Indonesia.

Data
Reanalysis observation data is used considering the absence or incompleteness of observation data at a location.The reanalyzed observation data used were Climate Hazard Group Infrared Precipitation and Stations (CHIRPS) for rainfall data and Agroclimatological indicators from the European Center for Medium-Range Weather Forecasts (ECMWF) for air temperature data.CHIRPS is rainfall data based on rain satellites combined with rainfall data from observations at rain stations.CHIRPS products are available in gridding format, starting in 1981, with daily temporal resolution and 0.05 O x 0.05 O spatial resolution.The results of the validation of CHIRPS rainfall data against observational rain data at rain stations show that CHIRPS has a good performance compared to similar rainfall data products [7].CHIRPS data can be accessed on the website: http://chg.geog.ucsb.edu/data/chirps/index.html.Agroclimatological indicators ECMWF ERA-5 is a reanalyzed dataset resulting from interpolation with a high-resolution model with a spatial resolution of 0.1 O x 0.1 O .Data on Agroclimatological indicators can be accessed at https://cds.climate.copernicus.eu/cdsapp#!/dataset/sisagrometeorologicalindicators?.

Calculation of the SPEI and SPI.
Drought indices that will be used in the study are SPEI and SPI.SPI is a drought index based on standardized rainfall, which is dividing rainfall over a certain period of time by its average value, then transforming it into a normal distribution using the gamma function to see the probability of rainfall that is below the average.SPI uses rainfall anomalies in relation to longterm normal conditions for several timescales [11].SPEI appears as a form of criticism for the SPI index, which does not consider other variables that can affect droughts, such as temperature, evapotranspiration, wind speed and groundwater capacity.In this study, SPEI and SPI were calculated using the SPEI package in the R program using a multi timescale over a period of 1, 2, 3, 6, 9, and 12 months.The calculation of seasonal SPEI and SPI is done by calculating SPEI and SPI with a time scale of 3 months, then the calculation in month t is calculated using rainfall data and evapotranspiration in months t-2, t-1, and t [12].

Trend test (MK test) and different test (T test).
Trend analysis is carried out by testing the significance of trends from threats that occur due to variability and climate change.The trend significance test was carried out using the Mann-Kendall (MK) test method.This test is widely used for trend detection in hydrological and climatological circuits [13,14].In this study, the MK test was used to estimate the trend of wet/dry changes during the season, namely the average SPEI and SPI indices in the months of DJF, MAM, JJA and SON.The trend will be significant if the value of | |≥ 1.96 at the 95% confidence level.Z value is a standard normal variable that represents the level of significance of a particular trend.A positive Z value indicates an increasing trend, and a negative value indicates a decreasing trend [6].The results of the calculation of the Z value are then mapped for the Selo watershed area.
The different test is used to see the effect of air temperature in creating wet/dry conditions in the Selo watershed by looking at the difference in the results of the dryness index calculation resulting from the SPEI and SPI calculations.If the different test shows that there is no difference between the drought index as a result of the SPEI and SPI relief, then the air temperature that is considered or used in calculating the drought index has no effect in creating wet/dry conditions in the Selo watershed.In this study, the difference test was performed using the t-test at a 95% confidence level.If the significance value (Sig. 2 tile) > 0.05, statistically, the average of the two data groups did not have a significant difference.

Monthly characterization of wet and sry condition in the Selo watershed
Characterization of wet/dry conditions, the first step is to make the average spatial data of the SPEI and SPI indices in the Selo watershed in the form of the cumulative frequency of dry and wet months from the extreme, severe, moderate categories and the number of extreme events (wet and dry) for each month.In the period 1981 -2020.The results are shown in Figures 3a -3c, namely, overall there were 77 wet months and 71 dry months that occurred in the Selo watershed during the last 40 years.
As seen in Figures 3a -3b, among the 77 wet months, there were 46 wet months in the moderate category, 22 wet months in the severe category, and 9 wet months in the extreme category.Meanwhile, of the 71 dry months, 37 dry months were included in the moderate category, 19 months were included in the severe dry episode category, and 15 dry months were in the extreme category.The results of statistical calculations show that the greatest chance of a moderate wet month occurring in June (6.3%)and December (5.4%), the greatest chance of a severe wet month occurring in August (3.5%)and March (3.0%), while the biggest chance for extreme wet months occurred in January and December with 1.6% and 1.1% chances, respectively.For the chance of a dry month, the dry month is moderate category, the biggest change is in March (4.8%) and September (3.8%), the biggest chance for a severe dry month is February (2.9%) and November (2.4%), while the biggest chance for extreme dry months occurred in June and October with 2.0% and 1.9% chance of occurrence, respectively.Figure 3c further it can be seen the cumulative number of each month opportunity for extreme conditions to occur, both extremely wet and dry.Cumulatively, the greatest opportunities for extreme conditions occurred in March (14.2%) and November (13.8%).The probability of occurrence of extreme conditions in the moderate category occurs in March (8.9%) and December (8.4%), for the medium category, the greatest opportunity occurs in March (5.0%) and August (4.6%),while the occurrence of extreme conditions in the extreme category of opportunities the largest occurred in June and October with a chance of 3.0% and 2.7%, respectively.Figure 4 is a cumulative frequency graph that illustrates the percentage of episodes of wet, dry and normal conditions with four different time periods, namely the period 1981 -1990, 1991 -2000, 2001 -2010 and the period 2011 -2020, in the period 1981 -1990 and 1991 -2000, dry conditions were more common than wet conditions.The opposite happened in the last two decades, when wet conditions were more common than dry conditions.It can also be seen that the highest percentage of dry conditions occurred in the decade 1991 -2000, while the highest percentage of wet conditions occurred in the period 2001 -2010.Meanwhile, the percentage of normal conditions mostly occurred in the period 1981 -1990.
In general, the processing and analysis of the above data show that the meteorological events that cause wet or dry conditions in the Selo watershed are rather complex, as indicated by the absence of a clear pattern of variability of wet and dry event episodes in the watershed.If only seen from the percentage of the number of extremely events with the highest probability of occurring in March and November, this only shows extreme wet conditions considering that March and November are the peak rainy seasons in the watershed area.Meanwhile, for dry condition episodes, which climatologically occurred in June, July and August, the accumulation of events was not seen, this shows the dominance of wet condition episodes when compared to dry condition episodes.Another complexity is indicated by the presence of different categories of dry and wet episodes that may have occurred in most months of the year.From the results of the analysis, it is also found that in the long term (decades) there is also no pattern of variability in wet and dry conditions in the study area, especially related to dry conditions that occurred in the period 1981 -2000 and wet conditions which occurred more in the last two decades (2001 -2020).

Seasonal spatial-temporal trends in wet/dry conditions in Selo watershed
To further study the temporal variability of wet and dry episodes in the Selo watershed, the MK test was applied to analyze the trend of wet/dry conditions each season (DJF, MAM, JJA and SON), where the seasonal wet/dry conditions were the average results.Each season from the calculation of the SPEI and SPI indices.Based on trend analysis using the MK test of the value of wet/dry conditions at extraction points in the Selo watershed (Figure 2), the Z value for each extraction point is obtained, the next step is to interpolate and map the Z value for each season in the Selo watershed, the results of which are shown in Figures 5a -5d.A positive Z value represents a trend towards wetter conditions, while a negative Z value indicates a trend towards drier conditions.
In Figures 5a -5d, it can be seen that all seasons in all watershed areas are characterized by a trend from dry conditions to wetter conditions, this is indicated by the Z value in all seasons in the region, which has a positive value.In the picture, it can also be seen the highest trend increase occurred in the DJF season and the lowest in the SON season, meaning that in the DJF season, the trend of change from dry conditions to wetter conditions was very significant, while in the SON season, the trend of changes from dry conditions to wetter conditions was not significant based on the MK trend test based on the Z value in these seasons.Based on the spatial distribution analysis, the trend of change from dry conditions to wetter conditions occurred in the upstream part of the west and a little area in the downstream part of the Selo watershed.Meanwhile, the central part of the Selo watershed experienced an increasing trend from dry conditions to lower wetter conditions.3.3.Effect of temperature on the occurance of wet/dry conditions SPEI and SPI are the most widely used drought indicators for meteorological drought estimation.In general, the SPI calculation process is almost the same as the SPEI calculation, as well as the index used to classify the intensity of wet/dry conditions, is also the same.However, there is a principle that distinguishes the calculation of the two drought indices.In calculating the drought index, SPI only uses rainfall data, while SPEI, in addition to rainfall, also considers air temperature to take into account evaporation under water equilibrium conditions.For this reason, in this study, a different test of the dryness index calculated by SPEI and SPI will be carried out to see whether air temperature has an effect or plays a role in creating wet/dry conditions in the Selo watershed in the context of global warming.
The test results at all data extraction points in the Selo watershed area showed no difference between the dryness index calculated by SPEI and SPI.This shows that the air temperature has no effect in creating wet/dry conditions.In other words, the wet/dry conditions in the Selo watershed are more influenced by the characteristics of the rainfall in the watershed.In contrast to the results of similar studies in the middle and high latitudes, where the effect of air temperature in creating wet/dry conditions is evident, there are differences in the results of the calculation of the drought index produced by SPEI and SPI [6].

Conclusion
This study uses the dryness index calculated by SPEI and SPI to conduct a quantitative investigation of the variability characteristics of meteorological wet/dry conditions in the Selo watershed, Tanah Datar Regency, for the period 1981 -2020.The results show that meteorological wet/dry episodes occur in the Selo watershed.Ithas interesting characteristics to study, including wet/dry conditions that do not have a clear pattern of variability in the long term, but it is slightly clear that there are variations in wet episodes that follow the pattern of the average amount of rainfall each month, this is not the case for dry episodes.Seasonal wet/dry conditions show a changing trend towards wetter conditions in all seasons.The most significant increase occurred in the DJF season and the least significant change occurred in the SON season.Spatially, these changes also occur in the entire Selo watershed area.The most significant increase occurred in the upstream and downstream of the watershed.For a longer time, it can be seen that in the last two decades, wet conditions were more dominant, replacing dry conditions that were more prominent in the previous two decades.The results of the analysis also found that wet/dry episodes with different categories can occur simultaneously, so it can result in more extreme wet/dry conditions to trigger severe droughts and floods.Other analysis results conclude that there is no influence of air temperature in creating wet/dry conditions in the Selo watershed area.These characteristics of wet/dry conditions are expected to be the basis for selecting agricultural adaptation strategies in the face of climate variability and change.

Figure 1 .
Figure 1.Research site: Batang Selo watershead.Source: SRTM imagery and Peta Rupa Bumi Indonesia.2.2 Data Reanalysis observation data is used considering the absence or incompleteness of observation data at a location.The reanalyzed observation data used were Climate Hazard Group Infrared Precipitation and Stations (CHIRPS) for rainfall data and Agroclimatological indicators from the European Center for Medium-Range Weather Forecasts (ECMWF) for air temperature data.CHIRPS is rainfall data based on rain satellites combined with rainfall data from observations at rain stations.CHIRPS products are available in gridding format, starting in 1981, with daily temporal resolution and 0.05 O x 0.05 O spatial resolution.The results of the validation of CHIRPS rainfall data against observational rain data at rain

2. 3 .
Data processing 2.3.1.Data extraction.The data extract points at the study site are evenly distributed both inside and outside the perimeter of the Selo watershed, as shown in Figure 2. The extract points are the midpoints of the grid box measuring 0.05O x 0.05O (~5.6 Km x ~5.6 Km).The results of the extract are time series data at a predetermined extraction point [8, 9, 10].One of the ways to extract data like this is by using the Grid Analysis and Display System (GrADS) software.The extracted data is then validated by using observational data from observations of rainfall and air temperature at the points closest to the extraction point.

Figure 2 .
Figure 2. Grid points in the Selo watershed.The center point of the grid (green dot) is the coordinate point for data extraction.The watershed background image is the daily CHIRPS rainfall data on February 13, 2020.

5 Figure 3 .
Figure 3. Monthly cumulative time: (a) dry episodes (b) wet episodes and (c) the number of wet/dry conditions from various categories of wet/dry conditions that occurred in the Selo watershed for the period 1981-2020.

Figure 4 .
Figure 4.The percentage of normal conditions and extreme events (wet/dry) per decade in the Selo watershed in the period 1981 -2020.

7 Figure 5 .
Figure 5.The spatial distribution pattern of seasonal wet/dry conditions in the Selo watershed for the period 1981 -2020.The figures in the legend are the Z values of the trend test results with the MK test.If the value | |≥ 1.96, the trend is at the 95% confidence level.