Introduction

background of RESEARCH PROJECT

Influence of the Boreal Summer Intraseasonal Oscillation on Monsoon Rainfall and Soil Moisture over the Western Ghats, India

Along with abundant precipitation, the India summer monsoon plays a critical role in the agriculture and hydroelectric power industry for the Indian subcontinent, but it is also strongly linked with the occurrence of flooding and rain-induced landslides by influencing soil moisture (Parthasarathy et al., 1987; Segoni et al., 2018). Due to its roles in social-economic activities and extreme hydro-climatic hazards, understanding the monsoon rainfall and its relationship with soil moisture is of great importance.

Modulated by the Boreal Summer Intraseasonal Oscillation (BSISO), the summer monsoon season (June to September, JJAS) over the India subcontinent is not characterized by a constant rainfall but by the alternate wet (active) and dry (break) spells lasting for two to three weeks (Webster et al., 1998). Of all the regions over India, Western Ghats, the mountain range along the western coast of India, receives the largest amount of rainfall (~3000 mm) in summer monsoon season as a result of the topographic response to the sustaining southwesterly monsoon flow (Romatschke and Jr., 2011). However, the impact of BSISO activity on rainfall characteristics and soil moisture response over the Western Ghats remains limited. With the help of the most recent data from two flux towers (Figure 1) on Western Ghats, there is a good opportunity to fill this research gap.

Aim: The research project aims to investigate how BSISO influences the rainfall variability and soil moisture response during summer monsoon seasons from 2016 to 2018 over the Western Ghats, India.

Figure 1. Location of flux towers at Dharwad and Berambadi with lines showing the monsoon onset (Credit: Andrew Turner).

Objective 1

To investigate the variability of rainfall and soil moisture under different phases of BSISO at Dharwad and Berambadi.

Objective 2

To assess the soil dry-down timescales following rainfall events at Dharwad and Berambadi.

Objective 3

To quantify the role of rainfall, soil moisture, and heat flux on soil dry-down events.

Methodology 1

Descriptive statistics and composite analysis

Methodology 2

Exponential decay modelling

Methodology 3

Correlation analysis

The technical part of my research project mainly involves data analysis and visualization, all of which are carried out with python programming. This website focuses on the technical information for achieving objective 1 and 2. The scripits associated with data processing and website development are hosted by both homepages.see system and Github. The homepage can be accessed via http://homepages.see.leeds.ac.uk/~ee18hz/ or https://haichen6622.github.io/WesternGhats/.

Initial Results

Figure 2. Rainfall anomaly over the Indian subcontinent during each BSISO phase (June-September 1999-2018)

The spatial rainfall anomaly was produced by the script of 'Rainfall_anomaly_20yr.py' with daily rainfall from (Tropical Rainfall Measuring Mission) TRMM 3B42 Multi-Satellite Precipitation Analysis dataset.

It gives information about the average daily rainfall anomalies over the Indian subcontinent during June-September in 1999-2018. Positive rainfall anomalies associated with BSISO begin from the southern part of India and Sri Lanka in phase 2 and then propagate northeastward to the Indian subcontinent and the Bay of Bengal during phases 3-6. After that, negative anomalies start to dominate the Indian subcontinent from south to north during phase 7, 8, 1 and 2.

In order to get the rainfall characteristics at Dharwad and Barembadi, average daily rainfall intensity during June - September 2016-2018 was calculated and aggregated to each phase of BSISO. The plots were generated by 'TRMM_daily.py' with 3-hourly TRMM 3B42 rainfall estimates.

Dharwad

Figure 3. Average daily rainfall at Dharwad during JJAS 2016-2018 when BSISO is acitive

Berambadi

Figure 4. Average daily rainfall at Berambadi during JJAS 2016-2018 when BSISO is acitive

The results show that average daily rainfall at Dharwad is higher during phase 4 - 6 and lower during phase 7, 8 and 1. The mean daily rainfall intensity during the summer monsoon season is around 8 mm/day. Different with Dharwad, the mean rainfall intensity at Berambadi is relatively lower with an average of 5 mm/day partly due to the stronger rain shadow effect, and the higher values appear in phase 3, 4 and 5. Considering the location of the flux towers, the patterns of daily rainfall intensity at each site are consistent with the propagation of spatial anomalies in BSISO phases.

The diurnal cycle analysis was also carried out with 3-hourly TRMM 3B42 rainfall dataset. The plots below were generated with 'TRMM_diurnal.py'.

Dharwad

Figure 5. Diurnal cycle of rainfall at Dharwad during June-September 2016-2018

Berambadi

Figure 6. Diurnal cycle of rainfall at Berambadi during June-September 2016-2018

At Dharwad, the mean diurnal cycle peaks in the late afternoon to early evening (14:30-17:30 IST) except for phase 5, 6 and 8, while at Berambadi, peak rainfall intensity tends to appear in the midnight in most of the phases. When BSISO activity is weak, the rainfall intensity becomes lower, and the diurnal cycle is insignificant.

Figure 7 and Figure 8 display the results of exponential fitting for soil dry-down events. Firstly, the dry-down events were extracted based on Integrated Multi-satellitE Retrievals for Global Precipitation Measurement (IMERG) rainfall dataset as well as in-situ measurements of soil moisture. The exponential fitting and plotting were achieved by the script of Drydown.py. Due to the incomplete flux tower data, the results shown here are only for examples.

Dharwad

Berambadi

Figure 7. Exponential fitting of dry-down events at Dharwad during JJAS 2016.

Figure 8. Exponential fitting of dry-down events at Berambadi during JJAS 2016

During the whole period of India summer monsoon season in 2016, there are only five soil dry-down events at Dharwad with e-folding timescales (τ) varied from 2.57 days to 24.71 days. The dry-down events in the middle of August were eliminated because of the irrigation. At Berambadi, seven dry-down events were identified for the period from June to the middle of August 2016. After that, no soil moisture data is available. The timescales for soil drying are between 0.61 days and 46.28 days.

Resources

Data processing of the research project is carried out with Python 2.7 in Spyder. All source codes attached in the following parts are hosted in Github Gist. Sample datasets and source codes can be downloaded from the corresponding link below.

Datasets

#	Dataset	Description	Unit	Link
1	TRMM_3B42_V7_Daily	Daily rainfall from TRMM dataset during 1999 - 2018	mm/day	Download
2	TRMM_3B42_V7_2016_2018_JJAS_3HR	3-hourly rainfall during June - September 2016 - 2018 from TRMM dataset	mm/h	Download
3	GPM_IMERG_WesternGhats_2016	30-min rainfall from IMERG dataset during June - September 2016	mm/h	Download
4	BSISO_25-90bpfil_pc_cdr_20yrs	BSISO index from 1999 to 2018	-	Download
5	BSISO_25-90bpfil_pc_cdr_JJAS_1618	BSISO index during June - September 2016 - 2018	-	Download
6	Dharwad_GapFilled	Soil moisture content at Dharwad	%	Download
7	Berambadi_GapFilled	Soil moisture content at Berambadi	%	Download

The TRMM 3B42 Multi-Satellite Precipitation Analysis datasets (0.25° spatial resolution and 3-hourly/daily temporal resolution) were extracted originally at https://pmm.nasa.gov/data-access/downloads/trmm (TRMM, 2011; TRMM, 2016), and GMP IMERG rainfall estimates (0.1° spatial resolution and 30-min temporal resolution) come from the https://pmm.nasa.gov/data-access/downloads/gpm (Huffman et al, 2019).

BSISO index is provided by the International Pacific Research Center (2019), which is available at http://iprc.soest.hawaii.edu/users/kazuyosh/Bimodal_ISO.html .

Soil moisture at Dharwad and Berambadi comes from the in-situ measurements of flux towers for the Joint project of INCOMPASS (Fletcher et al., 2018). The variables also include surface temperature, latent heat flux, sensible heat flux, near-surface wind speed and direction, etc. All have a temporal resolution of 30 min.

Scripts

#	Script	Description	Source Code
1	Rainfall_anomaly_20yr.py	To carry out spatial analysis of rainfall anomaly during each BSISO phase (June - September in 1999 - 2018)	Download
2	TRMM_daily.py	To calculate the average daily rainfall intensity during each BSISO phase at Dharwad and Barembadi	Download
3	TRMM_diurnal.py	To analyze the diurnal cycle of rainfall during each BSISO phase at Dharwad and Barembadi	Download
4	Drydown.py	To extract dry-down events and do exponential fitting for Dharwad and Barembadi	Download

Instructions:

1. To carry out data processing, firstly you need to download the required datasets that are indicated in the script part below.

2. Before running the script, you also need to set the working directory properly.

filepath_BSISO points to the folder where you store BSISO index.

filepath_rain points to the folder where you store TRMM/IMERG rainfall estimates.

filepath_plot points to the storage folder for plots saving.

filepath_process is the place to store processed data.

3. For running the scripts of TRMM_daily.py, TRMM_diurnal.py and Drydown.py, you need to indicate which site (Dharwad or Berambadi) you intend to analyse at the beginning of the script.

All the parts in python scripts required to be modified are marked with comments and the symbol of '--------------------------'.

Rainfall_anomaly_20yr.py

The spatial rainfall anomalies during each BSISO phase was carried out with the data in TRMM_3B42_V7_Daily folder and BSISO_25-90bpfil_pc_cdr_20yrs.txt. The rainfall anomalies at each grid point are calculated by:

This script was modified from the lecture materials of Visualising the Madden-Julian Oscillation in SOEE5710M Advanced Data Analysis and Visualisation for Environmental Applications (Birch, 2018). After getting the plots of rainfall anomaly during each BSISO phase, the animation in .gif format was created by GIFMaker.me.

TRMM_daily.py

The input rainfall data is stored in the folder of TRMM_3B42_V7_2016_2018_JJAS_3HR, and the referred BSISO index is the file of BSISO_25-90bpfil_pc_cdr_JJAS_1618.txt.

TRMM_diurnal.py

The input rainfall data for diurnal cycle analysis is TRMM_3HR_Dharwad.npy or TRMM_3HR_Berambadi.npy (click to download if you did not run TRMM_daily.py) that was produced when running TRMM_daily.py. The referred BSISO file is BSISO_25-90bpfil_pc_cdr_JJAS_1618.txt.

Drydown.py

For running Drydown.py, in-situ soil moisture (Dharwad_GapFilled.txt. or Berambadi_GapFilled.txt) and IMERG rainfall estimates (GPM_IMERG_WesternGhats_2016.npz) are required.

The first step is to extract the dry-down events from the entire study period. Combining the criteria defined by Shellito et al. (2016) and Salvia (2017), a dry-down event starts after the accumulated rainfall in the previous 24 hours greater than 5 mm and stops once 2 mm or above of subsequent rain accumulates or ends when there is soil moisture increase. In this script, dry-down events are firstly screened out automatically based on the IMERG rainfall estimates and then need to be manually checked to ensure that there is no unexpected increase in soil moisture (e.g. due to irrigation). After that, an exponential model will be employed to assess the timescale and magnitude of each dry-down event at the indicated site. The exponential decay function is shown below (Shellito et al., 2016):

where θ is soil moisture content at a certain depth (%), t is the time since the start of the dry-down event, A is the magnitude of soil dry-down (%), τ is the e-folding timescale, θ_f is the final soil moisture content (%) that is confined to be smaller than the lowest observed soil moisture during the dry-down event and equal to or higher than the lowest soil moisture of the site (Shellito et al., 2016).

According to McColl et al. (2018), only the dry-downs with the coefficient of determination from exponential fitting greater than 0.7 are selected for further analysis.

References

Birch, C. 2018. Visualising the Madden-Julian Oscillation. Lecture notes distributed in SOEE5710M Advanced Data Analysis and Visualisation for Environmental Applications. 12 December, University of Leeds.

Fletcher, J. K., Parker, D. J., Turner, A. G., Menon, A., Martin, G. M., Birch, C. E., Mitra, A. K., Mrudula, G., Hunt, K. M. R., Taylor, C. M., Houze, R. A., Brodzik, S. R. and Bhat, G. S. 2018. The dynamic and thermodynamic structure of the monsoon over southern India: New observations from the INCOMPASS IOP. Quarterly Journal of the Royal Meteorological Society. doi: 10.1002/qj.3439.

Huffman, G. J., Stocker, E. F., Bolvin, D. T., Nelkin, E. J. and Tan, J. 2019. GPM IMERG Final Precipitation L3 Half Hourly 0.1 degree x 0.1 degree V06. Goddard Earth Sciences Data and Information Services Center (GES DISC). [Online]. [Accessed 1 April 2019]. Available from: 10.5067/GPM/IMERG/3B-HH/06

International Pacific Research Center. 2019. Historical ISO index (CDR OLR). International Pacific Research Center. [Online]. [Accessed 5 April 2019]. Available from: http://iprc.soest.hawaii.edu/users/kazuyosh/Bimodal_ISO.html

McColl, K. A., Wang, W., Peng, B., Akbar, R., Short Gianotti, D. J., Lu, H., Pan, M. and Entekhabi, D. 2017. Global characterization of surface soil moisture drydowns. Geophysical Research Letters. 44(8), pp.3682-3690.

Parthasarathy, B., Sontakke, N. A., Monot, A. A. and Kothawale, D. R. 1987. Droughts/floods in the summer monsoon season over different meteorological subdivisions of India for the period 1871–1984. Journal of Climatology. 7(1), pp.57-70.

Romatschke, U. and Jr., R. A. H. 2011. Characteristics of Precipitating Convective Systems in the South Asian Monsoon. Journal of Hydrometeorology. 12(1), pp.3-26.

Salvia, M. 2017. Surface soil moisture spatio-temporal dynamics in southeastern South America observed by SMOS and modeled by ORCHIDEE [PowerPoint presentation]. 4th Satellite Soil Moisture Validation and Application Workshop, 20 September, Wien, Austria.

Segoni, S., Rosi, A., Lagomarsino, D., Fanti, R. and Casagli, N. 2018. Brief communication: Using averaged soil moisture estimates to improve the performances of a regional-scale landslide early warning system. Natural Hazards and Earth System Sciences. 18, pp.807-812.

Shellito, P. J., Small, E. E., Colliander, A., Bindlish, R., Cosh, M. H., Berg, A. A., Bosch, D. D., Caldwell, T. G., Goodrich, D. C., McNairn, H., Prueger, J. H., Starks, P. J., van der Velde, R. and Walker, J. P. 2016. SMAP soil moisture drying more rapid than observed in situ following rainfall events. Geophysical Research Letters. 43(15), pp.8068-8075.

Tropical Rainfall Measuring Mission (TRMM). 2011. TRMM (TMPA) Rainfall Estimate L3 3 hour 0.25 degree x 0.25 degree V7. Goddard Earth Sciences Data and Information Services Center (GES DISC). [Online]. [Accessed 10 March 2019]. Available from: 10.5067/TRMM/TMPA/3H/7

Tropical Rainfall Measuring Mission (TRMM). 2016. TRMM (TMPA) Precipitation L3 1 day 0.25 degree x 0.25 degree V7. Goddard Earth Sciences Data and Information Services Center (GES DISC). [Online]. [Accessed 20 February 2019]. Available from: 10.5067/TRMM/TMPA/DAY/7

Webster, P. J., Magaña, V. O., Palmer, T. N., Shukla, J., Tomas, R. A., Yanai, M. and Yasunari, T. 1998. Monsoons: Processes, predictability, and the prospects for prediction. Journal of Geophysical Research: Oceans. 103(C7), pp.14451-14510.

Technical Information

Introduction

background of RESEARCH PROJECT

Influence of the Boreal Summer Intraseasonal Oscillation on Monsoon Rainfall and Soil Moisture over the Western Ghats, India

Objective 1

Objective 2

Objective 3

Methodology 1

Methodology 2

Methodology 3

Initial Results

Dharwad

Berambadi

Dharwad

Berambadi

Dharwad

Berambadi

Resources

Datasets

Scripts

Rainfall_anomaly_20yr.py

Future Work

References

Haichen Zuo

MRes Climate Change and Atmospheric Science