flashfloods and weather forecasting

After Typhoon Sendong, rains may now be viewed as agents of disasters. But knowing rainfall forecast can actually save us from harm. We need only to know better our surroundings. This blog tries to explain how we should look at weather forecast with the end in view of better assessing risk vulnerability of our respective communities. I am emphasizing that we should not rely heavily on storm signals because they are all based on winds (please read switching from wind to rain). We should start listening to rainfall forecast and make sense of it. Later, we will use the Iligan case for illustration purposes.

So how rainfall forecast can save you? If you live near the river or at the footslopes of a mountain, you need to listen to the following information:

a) RAINFALL INTENSITY
b) SPEED AND DIRECTION OF MOVEMENT OF RAIN CLOUDS
c) SWATH OR DIAMETER OF THE CLOUDS

Our common mistake is we only focus on the eye of a typhoon and don't bother about how wide is the rain cloud system. More often we associate typhoon signal as the strength of rain. Which is why a lot of us would wonder why certain areas have signals in place but there are no rains. We sometimes ridicule PAGASA about this. But actually PAGASA based their warnings on wind strength. That is why we need to know the information about the rain and the cloud system behavior.

So here we are. Rainfall intensity gives us idea on how heavy is the rain. Usually it is expressed in mm/hr. So that if rain pours at 25 mm/hr for 8 hrs, then we expect to have received 200 mm. By the looks of it, you may say 200 mm or 20 cm is not that deep, right? About 2/3 of your child's ruler. But let us reserve that info later.

For Sendong, it was projected to have 10 to 25 mm/hr of intense rain. How do we predict the duration of the rain? This is where the swath of clouds come in. Usually, you can estimate the diameter by looking at the satellite imagery (usually red in color) or Kuya Kim will sometimes tell you this. For example, Sendong has been predicted to have a cloud diameter (swath) of 400 km. We should know also the speed of its movement. PAGASA pegged it at 24 km/hr. From here, we will compute how long will it take for a cloud system to pass a particular place from end to end. For Sendong, it would be 400 km / (24 km/hr) =  16.67 say 16 hrs of continous rain. If we expect 10 to 25 mm/hr for 16 hrs, the expected rain would be in the range of 160 mm to 400 mm. That means any particular place passed by Sendong potentially can receive 160 to 400 mm of rainfall depth. If I am not mistaken, CDO received 180 mm.

So what do we do with this 160 to 400 mm? This is were the concept of hydrological unit would come into play. Always remember that a certain place will respond differently given a certain estimated volume. This is where the role of PAGASA ends and where the community's vigilance begun. Let us use Iligan City as our case. I use Google Earth to render a 3D visualization.

     
The satellite image above was captured in July 2009. At first glance, Iligan is so pristine with all the lush green and mountains around it. These mountains in fact can shield the city from strong winds cause by typhoons. But beyond this imposing mountains lies a big container of water or watershed. Iligan was not hit by the winds, it was hit by waters that rushed in from the mountains. Now let me give you a tour what lies behind this Iligan beauty.

 
As you can see above, there are two watershed that drains to Iligan. One that has an area of approximately 65,000 hectares and one with around 7,800 hectares (estimated only through GE Path). The bigger watershed drains water from as far as Talakag in Bukidnon and Kapai and Tagoloan II in Lanao del Sur. The smaller watershed drains partly the towns of Tagoloan, Baloi, and Pantaran in Lanao del Norte. All of these watersheds drains to an area approximately 1,500 hectares within Iligan proper. It is like having two large buckets pouring their contents to a very small one.

Suppose 160 mm poured over these watersheds, we can use this to compute the flooding volume. We will use 160 mm because we do not know the exact observation there. Anyway, if others have the exact figure from any of the rain gauges in the city or on its watershed they can just follow the procedure herein and compute for themselves a precise figure.

Expected flooding volume = 160 mm (65,000 ha + 7,800 ha) = 0.16 m (728,000,000 m2)
that would be 116,480,000 m3 potential water volume. But of course, not all of these water will rush in. Since the watershed looks intact with lush forest, let us say only 50% (higher if without forest cover) of these waters flows into Iligan. So that would be around 58,240,000 m3. What does this figure mean for the Iliganos? Let us relate that to their low lying area (potential floodplains) which is around 1,500 hectares. Let us divide 58,240,000 m3 with 1,500 hectares: 58,240,000 m3/ 15,000,000 m2 = 3.88 m. This can be the estimated flood height. Enough to pile up three cars over each other. In short, if 160 mm of rain pours over their watersheds, Iliganos expects 3.88 m of floods (If the rainfall was 80 mm, then expected floods would be 1.94 m). Then the next to do is to stay away from the areas that can be inundated (can be reached by possible flood level). 

The foregoing discussion illustrates how rainfall forecast can be used to project potential flood level without sophisticated flood warning system. We deviate from the usual hydrological computation and use only the basic principle. This is the predicament of limited hydrological data in an area which is by the way expensive to conduct as well. While there are flood hazard maps in the area, these are without meaning in disaster preparedness unless linked with rainfall forecast. This heuristic approach is good for areas without the resources to buy flood warning systems. This is a crude way of localizing weather forecasts, something that makes sense in our localities.  

The question is how about your area? Do you have any clue what would happen if you expect 160 mm rainfall? Communities should also be aware of their sorroundings. How large is the watershed area? What is the shape of the river channel? Is its mouth constricted (similar to downspout) like that of Iligan? Sometimes, it pays to listen to the rhythm of the falling rain.   

(The figures above are estimates and is meant for discussion puposes only. Please also read: topography of flashfloods and typhoon warnings: switching from wind to rain)

topography of flashfloods

Amidst these news of flash floods, I explored the topography of the flooding sites namely that of Cagayan de Oro (CDO) after Typhoon Sendong (December 2011) and that of Davao City (last July 2011). I used 3D rendition via Google earth to see what is the current land use and topography. I do not have an exact knowledge about the actual land use but satellite imagery more or less can suggest actual land use. This however need to be verified on the ground. I do not know the names of the places so I just tag them as points A, B, C, D, etc.

Here are some screenshots for Cagayan de Oro City:

The wide swath of the typhoon can render runoff from points A, B, and D going to the vicinity of C which is approximately the position of Cagayan de Oro City.

I am not so sure if this is the flooded area, but based on videos on the internet, the features resembled that of the actual flooded area. Well, this rendition is just for discussion purposes on how land use can affect hydrology and how urban planning can render populace vulnerable to risk. Take for example this image above, it shows the topography like a trough. Unfortunately, most of the residences are within this trough.

Going back to our larger picture, we can actually see two watershed draining to the vicinity of C. We added point E because it seems D have its own pathway but still in the vicinity of C. Here, we illustrated that floodwaters coming from A and B are essentially choked at C (like a trough) and eventually flush out water (illustrated as a three pronged arrow). Floodwaters from D would eventually settle on the E floodplains. By the looks of it, it may not be as destructive as in point C.

So we have here (see image above) a situation that a large watershed drains to a small floodplains (marked CDO).

It is wide that point A to C is approximately 50 km and B to C is around 20 km. By the looks of it point A is already in Bukidnon.

Examining satellite images, we can see that land uses can be agricultural fields in point A and B. Looking closer at point A, it can be gleaned that it is of intensive agricultural land use as evidenced by its brown patches. That is also true with slopes at point D. Although at point D, we cannot discount possible logging activities there due the presence of forest patches. What is crucial here is to know (ground truthing) the actual land use activities on point A and B. Agricultural activities that need to be reviewed are vegetable and corn farming - considered to be erosive agricultural land practices. Are agricultural activities in those areas employ some sort of a soil and water conservation to arrest runoff?  If farmers do not consider this, dwellers in C will be forever at the mercy of floods. This large watershed may encompass several municipalities in Misamis Oriental and Bukidnon. Negotiating proper land use to protect CDO might be an uphill struggle but worth the try.

Zooming in on C, you will find several property development. Locally, if these subdivisions are constructed in such a way that drainage connects to the main river, it will add to the problem. This is particularly true if streets are paved and residences lawns are paved as well. I suggest that they employ some sort of water retention facilities (in agricultural engineering we called it bioretention) to contain storm water to at least delay the runoff towards to the river. Or design houses with rainwater harvesting or green roofs. For point C, they have no choice but to start building houses with second floor to adapt to the situation. Our forefathers are smarter than us. They built houses on stilts!

More or less, the Davao City incident has some pecularities with CDO case:

But unlike CDO, the Davao flashflood was triggered by localized rainfall on a smaller watershed. But the same mechanism is still in place and that is the presence of a "choked point". Looking at satellite imagery, the watershed were possibly plantation crops (banana, fruit trees). Unlike that of CDO, this is less erosive provided planted properly. What I mean by planted properly is that it should consider the contour and slope. But looking closely (please try Google Earth) some of this plantation crops was planted along straight lines. In this case, runoff is still high.

I could not discount the possibility of the contribution of property development or quarrying on top of the Matina floodplains. But I guess the biggest contribution comes from the agricultural plains in the north. A larger watershed is needed to impound such large volume of water to flush the floodplains. I mean flush not just flash.

I just hope that this will help explain some phenomena we experience nowadays. Hope that policy-makers will consider hydrology in crafting their respective land use plans. Again, these illustrations are meant for discussion purposes. We still need to verify on the ground the actual land use before we could definitely give conclusions.

Please read also: using rainfall to save yourself from floods and typhoon warnings: switching from wind to rain . If you think this is helpful, you are free to share.