Hello from Akash and I from South India.

We have just completed one week of our month-long field collection campaign in the Gundar River Basin ( see pin below).

After several travel days, we arrived in Bangalore early in the morning last Wednesday and Akash’s family kindly let us stay with them until our night train to Madurai in the evening. It was a short but very kind visit and when we left on the train, we had quite the sendoff from his family – even an uncle, aunt and cousin attended.

When we arrived in Madurai early Thursday, we immediately went to the Development of Human Action (DHAN) academy outside the city, which we stayed at until Sunday to get our bearings for the remainder of the month. We have now shifted to staying inside the city of Madurai, a bit closer to our field site and close to the DHAN main headquarters.

For those of you unfamiliar with our research, both Akash and I are interested in understanding catchment and basin-scale hydrology in the region, largely with remote sensing observations (that’s broad, I know). My work focuses more exclusively on tank irrigation structures (small reservoirs) and Akash is focused on the response of evapotranspiration to land use changes in the basin.

Hopefully I can convince him to do a guest blog soon!

During the next few weeks, I will at least blog weekly but I will try to do more. Some of you reading may be VERY familiar with my work because you followed previous blogs or work with me, but others may be new. So, I am going to give a brief background on DHAN, and my research. * The VERY is for my sister who graciously edits much of my writing.

DHAN Foundation is an NGO that works directly with communities to build their resilience and one of their key focus areas – among many – is the rehabilitation and usage of tanks across the basin. Tanks are small surface water structures that were developed millennia ago to store water for irrigation but the purpose of tanks today in a farming climate with dependence on groundwater pumping is not clear. Some studies show tanks are important for shallow aquifer recharge in the direct command area, while others show this is very location dependent and not always the case. The main area of my research is trying to gain more information about tanks (i.e., their usage, filling behavior, irrigation demand) and data-driven information from satellites provides an opportunistic avenue to offer a larger ‘view from above’ observation of tank state.

The extent and amount of work that DHAN does in the basin is quite impressive and we are fortunate to work directly with them, not only for the data collection, but to help narrow and focus our research questions to make actionable community outcomes. I am sitting in one of DHANs regional offices as I write this blog looking at a wall of poster printed tank cascades, labelled with visits and tasks showing the huge involvement they have in each of these rural Indian communities. On Friday I was discussing a tentative schedule with our field lead, and I was not expecting him to be available on the weekends. But he insisted he was available and said their work requires every Saturday and often on Sunday they provide education to the rural communities on water and the value of protecting it.

We have two main goals during the month. The first is to collect data to help our research – primary and/or secondary data; whatever data we can get our hands on given there is very limited digital information for the basin and much of the information on tanks is in hard copy records. And our second goal is to strengthen relationships with our NGO partner and allow a two-way exchange in knowledge sharing.

So for the data collection…

Bureaucracy in India and getting data is what I imagine being a hamster on a hamster wheel feels like. You continuously go in circles, take a break and start again. Akash has made great progress on drone permissions but is still being tossed around from office to office and it is still highly uncertain if we will get the permissions.  My presence makes it particularly difficult, but we are hopeful that with persistence, we will find a way. There is no direct channel to follow so it feels quite chaotic.

If we get the drone permissions, our goal is to collect several repeat images of the tanks to compare with satellite estimated tank surface water extent. Often very high-resolution imagery can suffice for validation of water extent from lower resolution sensors, but given this region is particularly cloudy, capturing cloud-free and repeat satellite observations is unlikely. So, we have been moving ahead with our plans since arriving on Thursday as if we will have the permissions in hand soon.

After having a team meeting with DHAN on Saturday and presenting our goals for the month, Akash set out (Again!) to get drone permissions and I visited our potential collection site with Elamugil, our field lead with DHAN. I easily had the more fun task for the day, but Akash took one for the team – Thanks Akash!

Prior to our arrival in India, I found two groups of reservoirs that appeared full, so these are the tanks we wanted to QA/QC to ensure they were accessible for the field team and contained enough water. We set out early in the morning to complete visits to all 13 tanks by noon. I was quite nervous that the tanks wouldn’t have much water (which would be unfortunate for my measurements) given I was here in Nov. 2019 and several tanks only reached 50% capacity. Luckily for me but more importantly for the farmers, this years’ monsoon (both the southwest starting in July and the northeast starting in October) were above average leading to several tanks in the selected cascades to be filled to 90% capacity. We traversed the tank bunds of the 13 tanks and identified the most suitable tanks.

Given our daily coverage limit of 3 sq. km, it is not possible to collect UAV flights of all 13 tanks. To increase our sample of water boundaries, I chose to collect ~ 0.25 sq. km segments covering several tanks. In the sample, I also included three complete tank areas; these tanks were selected as our radar imagery showed clear water definition. This was not an easy task as tanks contain large amounts of vegetation, so I attempted to choose samples that have deep and shallow water, and various degrees of vegetation that are accessible for the team. Not all tanks are easy to survey either because they are heavily covered by woody vegetation limiting access or are not connected to a road.

A second dataset we are hoping to acquire with the help of Planet Labs is some SkySat images. A request has been made to task one of their SkySat satellites, which would collect 50 cm multispectral images of these two cascades. The collection will depend on cloud-free conditions during satellite overpass and the number of tasks in the queue, so image acquisition is not guaranteed but we are excited for this possibility.

On Sunday Akash and I met the drone team and did a test flight at a plantation area to mount and test the camera we brought. The Phantom 4 drone camera has red, green, blue channels but for water mapping, especially turbid water, a near-infrared channel is ideal. So, we purchased a MAPIR camera and are currently getting up to speed with mounting, collection, processing and stitching. We have had a few hiccups and are still trying to optimize the exposure setting but we are getting closer! Luckily the MAPIR team is extremely helpful and available to troubleshoot the calibration and processing of the data. See our team in the images!

On Monday, Akash and I also went our separate ways. He is on attempt day 4 of drone permissions at the local police station in our district and I am in the field/DHAN office.  The drone team is setting up the ground control points in the tank segments with a differential GPS and several rover stations. While I started in the field, I quickly realized my presence was not only unhelpful, but potentially made their job more difficult because the locals were extremely curious with my presence. So, I returned to the nearest town with a DHAN office to troubleshoot some data processing. However, the limited internet access made it very difficult.

Yesterday Akash and I both took a must needed rest day and just worked getting some research/ta-ing completing on our own time. Today, Akash is headed to the police station to work on permissions again and we are really hopeful that we will get some good news today. We keep trying to remind ourselves that this is the reality of fieldwork and even if the permissions come later than expected, any observations would be a huge help for the science.

We also brought a spectrometer with us but I will update and explain our science goals for this in the next blog.

Favorite thing so far (once again) is the amazing, amazing milk tea.