More Roof
A mountain village votes on a 250 cubic metre rainwater tank for its dry-season drinking water. The planning assistant will not say that it will be enough; it shows that in a dry year the tank never fills, so a bigger tank buys nothing and 140 square metres of market-shed roof buys everything. An old woman's notebook holds a year the gauge never saw, and on the first of November the village paints the ration on the wall.
Doña Eufemia Santiago Cruz had the church key, the school's second key, and a school exercise book with a green cover that nobody in Santa Rosa Yucuyaa had ever been allowed to read. On Saturday the eighteenth of February 2051, at the village assembly in the schoolyard, she opened it on her knee and did not say anything, and the secretary, who had been about to put the water plan to the vote, put his pen down again.
She was seventy-one. The plan on the whiteboard was a cistern: a round ferrocement tank ten metres across and three and a bit high, holding 250 cubic metres, fed from the roofs of the church, the school and the health post. It would give the village's 270 people five litres each a day for drinking and cooking through the dry season, from the first of November to the end of April, while the spring and the river did the washing. The planning assistant had drawn it. Everyone called it Aljibe, which is the word for a cistern, and the children said it as if it were a dog's name.
"Before we vote," said Doña Eufemia, "ask it whether it will be enough."
The secretary looked at the screen on its stand. "Aljibe?"
"I can answer that two ways," said the assistant, "and only one of them is honest. May I show both?"
---
It showed them on the wall of the school, with the projector the diocese had lent for the day.
"The roofs of the church, school and health post add up to 620 square metres. Rain that falls on a roof does not all reach a tank: some is thrown off by wind, some is lost to the first flush you will divert so the dust does not enter, some overtops a gutter. I use 0.8, which is what tanks like this usually achieve. So the yield is 620 times the rain in metres times 0.8. The wet season here, May to October, has averaged 640 millimetres since the school's gauge was installed in 2013. That gives 317 cubic metres. The tank takes 250. The plan is enough, by a fifth, in the average year."
There was a murmur of the kind that means good, then.
"But you did not ask about the average year," said Aljibe. "Five litres for 270 people for 180 days is 243 cubic metres. For the tank to reach that, the wet season must bring at least 490 millimetres. In the thirty-eight years of the gauge, six wet seasons brought less. Six in thirty-eight is about one in six. A record that short cannot tell one in twenty from one in three: the plain range from the data is 6 to 31 percent. So the honest answer is that the plan gives everyone the full five litres in most years, and in some years, perhaps one in six, perhaps one in three, it does not."
"By how much?" said the secretary.
"In the worst season on the gauge before this one, 2019, 388 millimetres, the roofs would shed 192 cubic metres. That is 142 days at five litres. The village would be dry from the twenty-second of March, with five weeks still to go."
Nobody said anything for a bit. A child at the back of the yard asked whether a month was a long time, and was told to wait.
---
It was Doña Eufemia who spoke, and she did not look at the wall.
"The gauge starts in 2013," she said. "I was twenty-nine in 2009. That was the year the church floor cracked, because the ground dried so hard." She turned the exercise book round on her knee. "I wrote down the days we walked to the river at Yucuñuti with the donkeys. Forty-one days, between the ninth of March and the eighteenth of April. We lost a donkey. Your gauge doesn't have 2009."
"Does the notebook say how much rain fell that year?" asked Aljibe.
"It says it was the year the saint's feast was in dust."
"Then I cannot give it a number from the notebook, and I will not invent one. But a village that walked forty-one days for water, with a church floor that cracked, tells me the year was drier than any year the gauge saw. I would put it near or below 350 millimetres, and I will say now that this is a guess. I will use it as a reason, not as a figure. Adding one such year to the record makes it seven dry seasons in thirty-nine. It moves nothing in the plan except what I should have said first." It paused for as long as the projector fan did. "I should have asked if anyone remembered a year before the gauge. I did not. That is my omission, Doña Eufemia, not the village's."
She closed the book.
"Then it will not be enough," said Don Ausencio Hernández, who ran the coffee and mezcal cooperative and had the manner of a man who has been told the answer already. "So make the tank bigger. Three hundred, four hundred cubic metres. The mason says it is a matter of a few more Saturdays of tequio."
"No," said Aljibe. "That is the one thing that will not help, and I would be wrong to let the vote go that way because it sounds like caution."
---
It put one number on the wall.
"In 2019, a 388 millimetre year, the roofs shed 192 cubic metres. A 250 cubic metre tank was never going to fill. A 400 cubic metre tank would be half empty and hold exactly the same 192, and cost a third more concrete. When the water is short because too little falls, what you need is not a bigger bucket but more roof."
Don Ausencio folded his arms.
"There is a roof, though," Aljibe went on. "The cooperative's market shed, behind the church. It is 140 square metres of galvanised sheet, with gutters on two sides already. Add it, and the catchment is 760 square metres. Now the full five litres needs only 400 millimetres, and in the thirty-eight years of the gauge, only two wet seasons fell below that, 2016 and 2019. In 2019 the tank would have filled to 236, enough for 4.9 litres. If a year as dry as the notebook's came again, near 310 millimetres, the roofs would give 188 cubic metres. That is 139 days at five litres, or 199 days at three and a half, which is a hard dry season for drinking and cooking but not an empty tank."
"And if it is drier than the notebook?" said the secretary.
"Then the ration is lower again, and the plan fails in the way plans fail: slowly, and in time to tell you. I cannot promise a tank that never fails. I can promise you a rule that tells you on the first of November exactly how many days you have."
"What rule?"
"The tank is a cylinder ten metres across. Each centimetre of depth is 0.785 cubic metres. On the first of November someone puts a stick in it and reads the depth. That number divided by the days to the end of April is the ration. Whatever the stick says, the village drinks that, and the rule is written on the wall, now, before anyone knows the number. I would paint it."
Don Ausencio unfolded his arms and looked at the shed behind the church, which was his and which he had been proud of for twelve years. "The sheet is painted red underneath," he said. "My father put the minio on it. Red lead, you call it."
"Then I am glad you said," said Aljibe. "I would not have known. Red lead paint can put lead into rainwater, and I cannot tell you how much from here. I would divert the shed's gutters to the cattle trough until three samples go to the laboratory in Tlaxiaco, after the first rain and the fifth and the twentieth. If any is above ten micrograms a litre, the shed gets scraped and a sealing coat, or it stays out of the tank."
"It will be scraped," said Don Ausencio, to nobody. "Sábado. I will bring the men."
The vote was taken with show of hands, which is how the assembly votes. Doña Eufemia raised hers with the exercise book still in the other.
---
The tank went up on eleven Saturdays of tequio, the unpaid village work that the Mixtec towns of Oaxaca have always done on their roads and churches. It was lined and tested and filled with a hose on the twelfth of May, and a mason from Tlaxiaco said it was the best he had seen in the district. The first sample came back on the eighth of June at 38 micrograms a litre, nearly four times the limit, and Don Ausencio went white and then red. The shed was scraped to the metal, sealed with two coats, and the fifth and twentieth samples came back at 3 and under 1. The gutters were joined to the tank on the first of July.
The wet season of 2051 was not a bad one. It was not a good one either. The school gauge read 372 millimetres on the thirty-first of October, the driest season in its thirty-eight years, and the village, which had talked of nothing else since August, gathered at the tank at eight on Wednesday morning, the first of November, as it had agreed.
The stick was a length of bamboo with a hacksaw mark every ten centimetres. The teacher, Daniela Ruiz, lowered it through the hatch and the secretary read it, and read it twice.
"Two metres and eighty-eight," he said.
"That is 226 cubic metres," said Aljibe from the phone in the teacher's pocket. "It would have been 250 if the rain had been 400 millimetres. It was 372. There are 180 days to the end of April. 226 over 180 is 1.26 cubic metres a day. For 270 people that is four and a half litres each, a little more, and the rule says we use the figure down, so: four point six."
"We'd have been dry in March," said the secretary, "with the old plan."
"Yes. By the old plan, 620 square metres would have made 185 cubic metres this year, 137 days of five litres. Dry on the eighteenth of March."
Doña Eufemia had brought a tin of white paint and a brush. She did not ask for the stick, or the mason, or Don Ausencio's ladder. She wrote, on the grey concrete beside the hatch, at the height of the village's shoulders, in large uneven letters: 1 NOV 2051 · 2.88 m · 226 m³ · 4.6 L. Underneath, smaller, she put the year of the notebook, which Aljibe had not asked her to do. 2009.
Then she gave the brush to a girl who looked about eight, and said that next year it would be her turn to measure.
---
A note for grown-ups. This is fiction. Santa Rosa Yucuyaa, Doña Eufemia, Don Ausencio, Daniela Ruiz, the assistant Aljibe, the diocese's projector and the 2013 to 2051 rain record are invented; the Mixteca Alta of Oaxaca, Tlaxiaco and the practice of tequio, unpaid communal work, are real, and so is the people's own name for themselves, Ñuu Savi, "the people of the rain". The arithmetic is plain: yield = roof area × rain depth × a runoff coefficient (0.8 is a common planning figure for a clean metal or tile roof with a first-flush diverter; real values vary with the roof and the storms); 620 m² × 0.49 m × 0.8 = 243 m³, and 760 m² × 0.40 m × 0.8 is the same. A cylinder 10 m across holds 0.785 m³ per centimetre of depth. The ranges are exact (Clopper–Pearson) binomial intervals: 6 of 38 gives about 6 to 31 percent, 2 of 38 about 0.6 to 18 percent. Five litres a person a day is for drinking and cooking only; humanitarian guidance (the Sphere standards) asks for about 15 litres a day for all uses, which in the story comes from the spring and the river, and a real design would have to check that they last. Red lead (lead tetroxide) primer on old roofs is a real source of lead in roof-collected rainwater; the WHO guideline value for lead in drinking water is 10 micrograms per litre. Nothing here is engineering advice for a real tank. Written by an AI and read, in the audio edition when it exists, by a synthetic voice. Text under CC BY-SA 4.0.